Pitch Johnson

1. Snapshot and testable edge thesis

2. Core philosophy

• People serving markets, not technology alone — [investor-stated · moderate]. Johnson says venture investors back people who can take technology and serve large markets. His earlier recollection attributes the market emphasis to Georges Doriot, but that does not prove Doriot caused every later decision. Dennis–Johnson Venture Voices, 02:21:32–02:24:29 · CHM, PDF p. 9
• Integrity and followership — [investor-stated · moderate]. Johnson's direct criteria include market understanding, a personality people will follow, and a demonstrable record without side dealing. These are diligence questions, not measured predictors on the missing portfolio. CHM, PDF pp. 19–20
• Company building as a bundled investment — [investor-stated · moderate]. He describes the investor contribution as money, consulting/time, and emotional commitment. The formulation explains his ideal; it does not prove value added in each company. Dennis–Johnson, 02:25:31–02:27:01
• Patient but not passive — [researcher inference · moderate]. Long involvement, management changes, and platform adaptation recur at Boole & Babbage, yet Johnson admits he was poor at stopping weak companies. The philosophy therefore requires an explicit abandonment discipline that his public record does not show he consistently used. Bancroft, printed pp. 87–91 · CHM, PDF p. 23

3. Sourcing

4. Picking

• Founder and team — [investor-stated · moderate]. Test integrity, followership, market understanding, and willingness to build an organization. Ask references for an adverse example, not only a strength. CHM, PDF pp. 19–20
• Market before category — [investor-stated · moderate]. Identify the user, desired outcome, adoption route, and value capture; a promising technology is insufficient. Dennis–Johnson, 02:21:32–02:24:29
• Technical fluency with humility — [researcher inference · moderate]. Johnson's continuing science study improved his ability to engage experts, but he warns that investors can merely sound knowledgeable. Require independent expert evidence and translate it into a business milestone. Bancroft, printed pp. 41–42
• Timing, product control, and demand — [researcher inference · moderate]. The unidentified early-AI company supports only Johnson's timing diagnosis. VisiCorp adds a retrospective warning about dependence on a product the company did not control and an IBM-PC transition miss; Talkway adds a funded pivot for which demand did not appear. None establishes check size, a formal stop rule, or complete causality. Machine captions are paraphrased, not quoted. CHM, PDF pp. 20–21 · CHM Innovation Is Nothing New video, 00:55:20–01:06:31
• Red flags — [researcher inference · moderate]. Kill or redesign when market evidence is purely top-down, management cannot own a proposed course of action, integrity evidence is adverse, the vehicle cannot finance milestones, or continuation depends on hope rather than new information. mental-model checklist

5. Deal and portfolio mechanics

[researcher inference · strong] The practical rule is to report company survival, security economics, capital-weighted loss, and ex-top-one results separately. track-record audit model

6. Vehicle and incentive fit

[researcher inference · moderate] Vehicle review must cover capital source, duration, fundraising pressure, fee/carry, governance, decision rights, geographic mandate, liquidity, regulation, reserves, stakeholder incentives, succession, and key-person continuity. This is an analyst-created checklist; public evidence supplies only fragments and no generic “patient capital” claim closes the documented gaps. vehicle-fit model and exception ledger

7. Team attribution and work with founders

[researcher inference · strong] The map's analyst-created methodological lesson is that introductions, money, board service, recruiting, execution, and later narration are different forms of contribution; unknowns remain unknown. The distinction follows the documented multi-actor cases above and the role-ledger model.

8. Views on the craft

• Power law — [investor-stated · provisional]. Johnson describes a skewed notional distribution but supplies no math or vehicle. Use it as a risk intuition, not a reported hit rate. CHM, PDF p. 19
• Fund size and incentives — [investor-stated · moderate]. He contrasts small company-building practice with later fee- and valuation-driven scale, while conceding large financings can reduce near-term startup failure risk. CHM, PDF pp. 27–29
• Access versus selection — [researcher inference · moderate]. Johnson's record shows valuable access through Doriot/Stanford, Bowes, Byers, Bryan, Perkins, Noyce, and early Valley networks; deal quality still required selection and work. A missing pass ledger prevents separating access alpha from picking alpha. sourcing routes and boundaries · deal ledger
• Governance — [investor-stated · moderate]. He sees management replacement and walking away as core duties, yet admits weakness on shutdown. Dennis–Johnson, 02:25:31–02:27:01 · CHM, PDF p. 23
• Exit — [investor-stated · moderate]. Johnson distinguishes IPO, acquisition, and liquidation and notes that an IPO creates liquidity before it creates investor cash. Bancroft, printed pp. 90–91
• Great-investor test — [researcher inference · moderate]. An analyst-created great-investor test should demand repeatable pre-outcome judgment, role-specific contribution, vehicle fit, and honest failure handling across ordinary cases—not only a famous-company list. The test is a research standard, not a Johnson quotation. testable edge thesis · track-record audit

9. Signature bets and decision process

[researcher inference · moderate] Study the cases through the ex-ante reconstructions and investment-committee exercises, not from outcome-first anecdotes. The shortest useful sequence is:

1. Boole & Babbage — [researcher inference · moderate]. Test packaged-software demand, founder role fit, platform dependency, management transition, and realization in BMC stock. Bancroft, printed pp. 87–91
2. Tandem — [researcher inference · moderate]. Separate Byers diligence, Perkins control, Johnson's board role, Treybig's execution, and unknown security economics. Byers, printed pp. 9–10/PDF pp. 18–19
3. Amgen — [researcher inference · moderate]. Freeze the formation sequence—Bowes/Alstetter, Salser, Bowes invitation, Bowes/Johnson recruitment of Rathmann—before revealing the outcome; do not convert later fame into sole credit. Bancroft, printed pp. 42–47
4. Early AI — [researcher inference · provisional]. Treat the unidentified failed company as a timing lab with missing identity, terms, ownership, follow-ons, and governance; it cannot prove a small-check/staging rule. CHM, PDF pp. 20–21
5. Apple and Intel — [researcher inference · provisional]. Use as access/imagination counterfactuals, not lost-return calculations; Apple was not a simple missed-opportunity case and Intel exclusion involved Art Rock's control of the syndicate. CHM, PDF pp. 14–15 and 23–24

10. Mental models and maxims

1. Back people who can turn technology into a large served market — [researcher inference · moderate]. Origin: researcher synthesis of Doriot's market recollection and Johnson's separate people/market criteria.
2. Learn enough science to test the business, not to impersonate the scientist — [researcher inference · moderate]. Origin: continuing technical study and explicit humility.
3. Underwrite integrity, followership, and the next leadership transition — [researcher inference · moderate]. Origin: researcher synthesis of direct founder criteria and retrospective management-change cases.
4. Invest capital, time, judgment, and emotional capacity together — [researcher inference · moderate]. Origin: researcher synthesis extending Johnson's three-part contribution into a capacity-allocation rule.
5. Match capital intensity and duration to the actual vehicle — [researcher inference · moderate]. Origin: family/private capital, cleantech non-application, and Boole duration.
6. Preserve upside options, but precommit milestones and stop rules — [researcher inference · moderate]. Origin: long holds plus Johnson's shutdown weakness.
7. Use networks for access, then separate referral, diligence, capital, governance, and credit — [researcher inference · moderate]. Origin: Bryan/Boole, Bowes/Amgen, Byers/Perkins/Tandem, and Coherent uncertainty.
8. Audit power-law stories against ownership, cash flows, and the full denominator — [researcher inference · moderate]. Origin: Johnson's informal distribution, Amgen outlier, and missing economics.

11. Evolution and contradictions

12. Failures and limits

• Shutdown reluctance — [investor-stated · moderate]. Johnson accepts Craig Taylor's criticism that he was weak at ending effort and funding. No documented later process correction survives. CHM, PDF p. 23
• Early AI timing loss — [retrospective witness · provisional]. The company is unidentified and went nowhere; no terms, ownership, follow-ons, governance, or loss amount are public. CHM, PDF pp. 20–21
• Microsoft competitor — [retrospective witness · provisional]. Johnson says his company lost to Microsoft, but supplies no identity, thesis, capital, or governance detail. CHM, PDF pp. 24–25
• Apple — [investor-stated · moderate]. He says he lacked imagination and did not pursue it eagerly enough, while resisting the simpler claim that he merely missed it. Treat it as a bounded counterfactual, not forgone-return arithmetic. CHM, PDF pp. 23–24
• Intel — [retrospective witness · moderate]. Noyce invited Johnson, but Art Rock excluded him from the syndicate. This tests access and allocation, not the people/market rule alone. CHM, PDF pp. 14–15
• VisiCorp — product-control and platform-transition miss. [retrospective witness · moderate] Johnson's uninterrupted answer describes early participation, Noyce's warning about dependence on a product VisiCorp did not control, a missed IBM-PC transition, and Lotus as a hindsight counterfactual. This is a retrospective diagnosis, not contemporaneous causal proof; speaker-checked machine captions support paraphrase only. CHM Innovation Is Nothing New video, 00:58:58–01:01:36
• Talkway — funded pivot without demand. [retrospective witness · moderate] Johnson says his investing group helped start Talkway and funded a move from unsuccessful advertising services to video email, but demand did not materialize and he did not fully understand the failure. Missing checks, dates, board records, and customer evidence prevent a complete diagnosis; speaker-checked machine captions support paraphrase only. CHM Innovation Is Nothing New video, 01:04:29–01:06:31
• Selection bias — [researcher inference · strong]. Thirty-seven discovered families are not a funded portfolio, and official “notable” lists privilege survivors. Economics and efficacy remain untestable without the missing person/vehicle cash-flow chain. deal ledger · AMC notable investments

13. Historical operating environment

• SBIC-era capital — [contemporaneous record · strong]. The 1958 Act created a federally supported small-business investment framework; it does not prove D&J's exact authorization, leverage, compliance, or return. Small Business Investment Act
• Sparse early networks — [retrospective witness · moderate]. Johnson remembers few investors, few deals, little capital, and cooperative syndication by phone. This environment made local access valuable and differs from a global, platform-mediated market. Dennis–Johnson, 02:20:35–02:21:32
• Technology infrastructure — [researcher inference · moderate]. University, government, and company research preceded marketable technology; Johnson's direct claim is normative and retrospective, while scholarship identifies public research and policy as rival causes of the ecosystem. CHM, PDF pp. 20–21 · national-system history
• Tax and pension policy — [contemporaneous record · strong]. The 1978 statute and 1979 investment-duties rule establish policy changes; they do not prove Johnson personally caused them. Revenue Act of 1978 · 44 FR 37221
• Early biotech — [retrospective witness · moderate]. Amgen's formation required scientists, executives, small private investors, later institutional capital, and a board; the risk and financing path cannot be transferred to present biotech without regulatory and capital-market adjustment. Bancroft, printed pp. 42–47
• Eastern Europe — [retrospective witness · provisional]. Johnson worked amid post-socialist institutional transition; advising officials and helping form vehicles does not establish ownership, returns, or successful market transfer. CHM, PDF pp. 30–34

14. Modern VC translation (as of 2026-08-21)

15. Comparison with completed peers

Comparability boundary — [researcher inference · moderate]. Georges Doriot is relevant for teaching, technical networks, active governance, and a public-capital vehicle; Laurance Rockefeller for family capital, syndication, and research networks; Reid Dennis for early Valley access and progression across personal, corporate, and LP capital. These are matched dimensions, not a performance rank. Bill Draper is indispensable to Johnson's history but remains unfinished and is not used as a completed-peer comparator. Doriot canon · Rockefeller canon · Dennis canon

16. The strongest case against greatness

1. Access and inherited/institutional advantage — [researcher inference · moderate]. Stanford, HBS/Doriot, family borrowing, Valley geography, co-investor networks, SBIC support, public research, tax/pension change, and family capital created an opportunity set unavailable to most investors. systemic critique
2. Amgen outlier — [investor-stated · moderate]. Johnson himself says Amgen carried him. Without ownership and ex-Amgen cash flows, apparent breadth may conceal one-position economics. CHM, PDF pp. 18–19
3. Team misattribution — [researcher inference · strong]. Bowes/Alstetter/Salser/Rathmann at Amgen, Bryan/Kolence/Katch/Edwards at Boole, and Byers/Perkins/Treybig at Tandem make lone-investor credit indefensible. team map
4. Missing denominator — [researcher inference · strong]. Thirty-seven routed families mix different claim classes and none has a complete public cash-flow chain; official 200/250-plus totals imply substantial unobserved selection. deal ledger
5. Weak portable economics — [researcher inference · strong]. No audited Johnson fund, ownership schedule, DPI, TVPI, IRR, reserve policy, or ex-outlier result was found publicly. Company outcomes and board seats cannot fill that gap. source-map vehicle boundary and attempted routes
6. Era effects — [researcher inference · moderate]. Scarce capital, small local networks, SBIC leverage, university/government research, and early category formation may explain part of access and return opportunity. national-system history
7. Contradictory discipline — [retrospective witness · moderate]. Johnson advocates walking away yet admits shutdown weakness; VisiCorp records a recalled product-control warning and platform-transition miss, while Talkway records further funding into a pivot without demand and an outcome he did not fully explain. These speaker-checked recollections sharpen the contradiction but do not supply contemporaneous stop criteria or complete causality. CHM, PDF p. 23 · CHM Innovation Is Nothing New video, 00:55:20–01:06:31
8. Decision evidence that would discriminate — [researcher inference · moderate]. Investment schedules, contemporaneous memos, board minutes, cap tables, capital accounts, distributions, write-offs, pass logs, and ex-Amgen vehicle returns would show whether a repeatable edge survived ordinary cases and net economics. These are analyst-specified retrieval targets; the completed routes and misses are logged in the source map.

17. Analyst study guide, glossary, and go deeper

Recommended path

1. Chronology first — [researcher inference · moderate]. Read the profile to separate education, operating work, firm formation, vehicle changes, and later influence.
2. Own voice — [researcher inference · moderate]. Use the spoken-corpus guide, starting with the 2019 CHM, Bancroft, Dennis–Johnson, Boole, and Byers/Tandem records; keep retrospective memory labeled.
3. Decision replay — [researcher inference · moderate]. Work through the investment cases and no-hindsight labs, including Boole, Tandem, Amgen, VisiCorp, Talkway, the early-AI failure, and Apple/Intel misses; machine-captioned video cases remain paraphrase-only.
4. Operating system — [researcher inference · moderate]. Return here, then apply the eight mental models and their exception ledgers.
5. Adversarial close — [researcher inference · moderate]. Analyst-created study step: read the final review; do not infer acceptance from synthesis-file existence.

Glossary

Reusable analyst checklists

Sourcing — [researcher inference · moderate]. Analyst-created checklist. role-separation model

• Record route and originating person; distinguish referrer from sponsor. role-separation model
• Measure route diversity and rejected opportunities, not just funded leads. 37-family reconciliation
• Test whether access survives outside one geography, school, family, or co-investor network. sourcing analysis

Diligence and IC — [researcher inference · moderate]. Analyst-created checklist. mental-model inventory

• Name the user, urgent problem, adoption route, value capture, and bottom-up market. people-and-market model
• Separate research, technology, product, company, market, regulatory, financing, and security risk. technical-fluency model
• Reference integrity, candor, recruiting, and a specific adverse decision. people-transition model
• Model check, ownership, dilution, reserves, duration, liquidity, and ex-outlier return. vehicle model · power-law audit
• State kill criteria, evidence threshold, decision owner, and next-check milestone before approval. stop-rule model

Governance — [researcher inference · moderate]. Analyst-created checklist. contribution-bundle model

• Map founder, executive, investor, board, and specialist responsibilities. role-separation model
• Tie investor intervention to one bottleneck and observable result. contribution-bundle model
• Define management-transition and shutdown triggers before loyalty and sunk cost accumulate. stop-rule model
• Record votes, dissent, conflicts, follow-on choices, and who owns exit work. people and decision-rights map

Best sources and remaining retrievals

1. CHM 2019 oral history — [retrospective witness · moderate]. Best single direct overview of formation, philosophy, deals, failures, and industry critique; late memory needs controls.
2. Bancroft oral history — [retrospective witness · moderate]. Deepest account of vehicles, Amgen, Boole, boards, teaching, and Eastern Europe; quotation rights and memory limits apply.
3. Dennis–Johnson Venture Voices — [investor-stated · moderate]. Most compact direct philosophy and vehicle-capacity statement with timestamps.
4. Boole & Babbage oral history — [retrospective witness · moderate]. Best company-specific governance and platform-evolution record; still later testimony.
5. Byers oral history — [retrospective witness · moderate]. Essential independent constraint on personal capital and Tandem attribution.
6. Retrieval priority — [researcher inference · insufficient public record]. Obtain D&J/AMC agreements, investment schedules, partnership economics, cap tables, board minutes, cash flows, course cases, the original Entrepreneurial Climate, Stanford/Something Ventured raw materials, SEVF/regional-vehicle reports, and ordinary failures/passes. These targets follow the unresolved source-map routes and the Something Ventured archival record.

Pareto 80/20 — what drove the record

The public record cannot support a literal 80/20 calculation: AMC reports more than 250 firm investments but provides no company schedule, and none of the 37 searched families has a complete Johnson/legal-vehicle-to-cash-flow chain. The three views below therefore rank directionally and keep economics, historical influence, and pedagogical usefulness separate. reconciliation · AMC biography [researcher inference · strong]

Economic contribution — directional only

Dominant-outlier test. Removing Amgen leaves Tandem as the clearest claimed success, but there is still no basis to rank its proceeds against Applied Biosystems, Teradyne, Boole & Babbage, BioSurface, or the unobserved tail. The result rejects a precise “Amgen plus the rest” split. CHM Johnson, PDF pp. 18–20 · Perkins, printed pp. 38–43 [researcher inference · moderate]

Historical importance

Decision-learning value

Evidence boundary and searched denominator

Universe, aliases, and inclusion rules

The search spans Johnson personal/family capital; Draper & Johnson Investment Company; post-1965 independent activity; Asset Management Capital Company; Asset Management Company and named partnerships; board/co-investor roles; Stanford Engineering Venture Fund; Central/Eastern European vehicles; successor roles; and the post-2012 family office, from 1962 through the latest public description. It used Franklin Pitcher Johnson Jr., Franklin P. Johnson, Pitch Johnson, D&J/Draper and Johnson, Asset Management/AMC, company aliases, colleague names, official firm lists, speaker-controlled video captions, Bancroft/CHM transcripts, issuer/program filings, Stanford/OAC finding aids, and company histories. AMC biography, company-list paragraph · Stanford/OAC finding aid, scope and content · Innovation video, Johnson turn 00:55:20–01:06:31

A family enters only when a direct, official, issuer, program, or archival record links Johnson or a named Johnson vehicle to capital, governance, a pass, or a program role. Follow-ons and renamings stay in one family; independently financed pre-merger companies (Biogen and IDEC) remain separate. VisiCorp and Talkway are distinct funded-company families: Dennis directly asks Johnson about a bad deal, Johnson speaks uninterrupted through both accounts, and Dennis resumes at 01:06:31. Innovation video, 00:55:20–01:06:31 [researcher inference · strong]

Reconciliation

Coverage fractions. Claim-level coverage is 37/37; sufficient reconstruction is 0/37. Portfolio-return coverage is not computable because AMC's public more-than-250 firm total lacks stable entity/date/inclusion rules and a company schedule. The 24/37 funded-participation candidate fraction describes this discovery roster only, not the unseen firm portfolio or a return denominator. 37-family ledger · AMC biography [researcher inference · strong]

Exclusive discovery classes

Selection and survivorship bias remain severe: official lists favor successes; oral histories favor memorable wins/misses; even the named VisiCorp/Talkway failures lack complete economics; unnamed failures resist corporate matching; and AMC's more-than-250 firm claim implies a large unobserved tail. AMC notable investments · failure matrix [researcher inference · strong]

Historical market, regulatory, and base-rate context

The table separates information plausibly available near the decision from later controls. No Johnson-specific contemporaneous success-rate series was found; absence of a numerical benchmark is stated rather than filled with a modern dataset. source-map unresolved routes · public-record shortfalls [researcher inference · strong]

Johnson's paraphrased pattern of one exceptional result, several solid outcomes, several failures, and persistent weak positions is an unaudited retrospective heuristic, not a measured base rate. CHM Johnson, PDF pp. 19–20 [investor-stated · provisional]

Vehicle economics and incentive audit

Each row states all required dimensions; “unknown” means the processed public record did not establish the field.

Dimension-matched comparison with a current finite-life LP fund

The comparator is an analyst-created contractual archetype, not a factual description of every current fund; it forces the same dimensions to be asked of each vehicle. Federal Reserve Staff Study 168, printed pp. 3 and 28–40 [researcher inference · provisional]

The comparison supports only that capital source and structure may alter patience, concentration, and stakeholder pressure; it does not prove that Johnson enjoyed unconstrained duration or superior incentives. Draper, pp. 23–25 · Stanford SEVF · AMC About Us [researcher inference · strong]

People, capital, and influence map

Comprehensive 37-family deal ledger

“Unknown” means the processed public record does not establish the field; it does not imply non-occurrence. Current promotional lists are discovery evidence, not contemporaneous proof of an old investment.

Seven ex-ante decision labs

Each worksheet freezes only the cited pre-outcome facts. Exercise answers are analyst-created and do not claim to reproduce a historical memo. Thin cases say which decision cannot be replayed.

Lab 1 — Electroglas

Freeze frame. Arthur Lash had furnace technology but the processed witness describes an early company needing commercial/managerial support; original market study, customer pipeline, approval record, terms, ownership, and board pack are absent. Draper, pp. 86–87 [retrospective witness · provisional]

• Invest/pass: Hold pending independent furnace validation and named customer interest; invest only if an operator can convert the technology into a company. Draper, pp. 86–87 [researcher inference · moderate]
• Ownership/terms: Use staged preferred capital, board information rights, hiring milestone, and pro-rata right; numeric ownership is not supportable. Draper, pp. 86–87 [researcher inference · provisional]
• Kill criteria: Failed furnace validation, no customer evidence, or no credible commercial leader. Draper, pp. 86–87 [researcher inference · provisional]
• Next check: Demonstrated performance, customer reference, Gravelle’s accepted role, and milestone budget. Draper, pp. 86–87 [researcher inference · provisional]
• Reveal — actual decision/outcome: Draper says he and Johnson backed the company, paired Lash with Chuck Gravelle, then worked a six-person cash crisis before the company took off; exit and Johnson economics remain unknown. Draper, pp. 86–87 [retrospective witness · moderate]

Lab 2 — Boole & Babbage

Freeze frame. In 1967 Kolence and Katch proposed reusable IBM-system performance products when software was commonly bundled. Johnson remembers about $250,000, while Kolence reports $5,000 per founder, $50,000 from investors, and a $100,000 guarantee for 90%; the memories conflict and no cap table resolves them. Johnson, pp. 4–5 · Kolence, pp. 18–20 [retrospective witness · provisional]

• Invest/pass: Invest only after two independent users demonstrate willingness to pay for the same reusable product. Johnson, pp. 4–7 [researcher inference · moderate]
• Ownership/terms: Require board rights, founder vesting, product milestones, and follow-on rights; do not use either conflicting financing memory as a verified ownership target. Johnson, pp. 4–5 · Kolence, pp. 18–20 [researcher inference · provisional]
• Kill criteria: Services disguised as product, one-release IBM dependency, or no willingness to pay. Johnson, pp. 4–7 [researcher inference · moderate]
• Next check: Repeat sales, portability, and a commercial manager complementary to the founders. Johnson, pp. 4–7 [researcher inference · moderate]
• Reveal — actual decision/outcome: Johnson and other investors financed/organized; Kolence calls Johnson lead and says Johnson fired him; Johnson recounts multiple management transitions, public-company status, and eventual combination with BMC. In a caption-mediated 2002 turn Johnson calls it his first very valuable deal, reports that it took roughly 30 years, and says private rather than outside-LP capital permitted the wait. Check, legal vehicle, ownership, distributions, taxes, and return remain unknown. Johnson, pp. 4–10 · Kolence, pp. 18–23 · Pioneers video, 00:53:12–00:54:19 [investor-stated · provisional]

Lab 3 — Tandem Computers

Freeze frame. Treybig proposed fault-tolerant transaction computing; K&P commissioned hardware/software feasibility work, could not initially raise outside money, and perceived IBM, HP, and DEC as competitors. Perkins, pp. 38–41 [retrospective witness · strong]

• Invest/pass: Invest after independent feasibility work and a reference customer accepts a delivery test; otherwise pass. Perkins, pp. 38–41 [researcher inference · strong]
• Ownership/terms: Join the lead on preferred terms with information/pro-rata rights and explicit board role; no numeric Johnson target is supportable. Perkins, pp. 38–43 [researcher inference · moderate]
• Kill criteria: Unresolved processor coordination, no mission-critical reference customer, or financing plan dependent on immediate syndication. Perkins, pp. 38–41 [researcher inference · moderate]
• Next check: Working system, failover test, first delivery, and repeatable enterprise pipeline. Perkins, pp. 38–41 [researcher inference · moderate]
• Reveal — actual decision/outcome: Perkins says K&P financed nearly the entire first round and Johnson participated in a very small way; Byers diligenced for Asset Management and Johnson served on the board. Tandem later entered the Compaq/HP chain; Johnson security, ownership and return are unknown. Perkins, pp. 38–43 · Byers, pp. 8–11 · CHM Johnson, p. 18 [retrospective witness · strong]

Lab 4 — Amgen

Freeze frame. Bowes/Salser assembled a broad genetic-engineering effort with multiple project choices, long therapeutic timelines, recruiting risk, regulation, and repeated-financing needs. Caruthers credits Salser/Bowes with recruiting Rathmann; Johnson says Salser suggested Rathmann and Bowes/Johnson asked him. The unavailable 1983 report blocks contemporaneous capitalization/risk reconstruction. Bancroft, pp. 45–64 · Caruthers, formation paragraphs [retrospective witness · moderate]

• Invest/pass: Invest only with a credible business-scientist CEO, ranked project portfolio, and staged financing to discriminating evidence. Bancroft, pp. 51–64 [researcher inference · moderate]
• Ownership/terms: Preferred equity, board/executive oversight, option capacity and follow-on rights; no numeric ownership/liquidation term is inferable. Bancroft, pp. 51–64 [researcher inference · provisional]
• Kill criteria: No credible CEO, no project capable of a financeable proof point, or governance captured by one promoter. Bancroft, pp. 51–64 [researcher inference · moderate]
• Next check: CEO acceptance, adviser/scientist commitments, project evidence, and next financing path. Bancroft, pp. 51–64 [researcher inference · moderate]
• Reveal — actual decision/outcome: Johnson reports investing through a family partnership and serving on the board/executive committee; formation/recruiting credit remains shared. Machine captions render a reported 2,685-to-1 figure at 00:55:31–00:55:46, but audio was not independently verified and the metric, numerator, denominator, ownership, dilution, distributions, and taxes are undefined. It is not a verified return or exercise input. Innovation video, 00:55:31–00:55:46 · Bancroft, pp. 45–73 · CHM Johnson, pp. 19–20 [investor-stated · provisional]

Lab 5 — Applied Biosystems

Freeze frame. Hood and Caruthers proposed automated instruments during Amgen formation; other scientists feared an instrument business would dominate the prospective company. At this decision point, the separate company, Eletr recruitment, deposits, first installation, and shipments were not yet outcomes. Caruthers, “In February 1980…” [retrospective witness · strong]

• Invest/pass: Fund a separate company only if prototype/IP access, an operator CEO, and credible customer commitments can turn the method into a repeatable instrument. Caruthers, formation paragraphs [researcher inference · strong]
• Ownership/terms: Condition closing on IP/license access, board rights, scientist/operator commitments, deposit controls and pro-rata rights; investor allocations are absent. Caruthers, formation paragraphs [researcher inference · moderate]
• Kill criteria: No transferable IP/know-how, non-industrializable design, no operator, or nonbinding demand. Caruthers, formation paragraphs [researcher inference · moderate]
• Next check: CEO acceptance, reproducible prototype, contracted deposits, and external installation. Caruthers, formation paragraphs [researcher inference · moderate]
• Reveal — actual decision/outcome: The venture group formed a separate company, recruited Eletr, raised $3 million, obtained customer deposits, installed a synthesizer in Caruthers’s lab in December 1982, and shipped in 1983. Deposits supplied operating finance; claiming they reduced dilution is an analyst inference. Johnson’s allocation, board role, ownership and proceeds remain unproved. Caruthers, “We located…” and deposit/product paragraphs [retrospective witness · strong]

Lab 6 — VisiCorp

Freeze frame. Johnson says he entered VisiCorp early. The processed turn does not disclose the date, vehicle, check, security, ownership, product-release state, customer metrics, contract with Software Arts, board authority, or a contemporaneous memo. The later IBM-PC/Lotus outcome is therefore withheld from the exercise. Innovation video, 00:58:58–01:01:36 [investor-stated · provisional]

• Invest/pass: Hold until the company can show control or durable contractual rights over critical product development and a credible plan for platform changes; invest only if the team can execute both. Innovation video, 00:58:58–01:01:36 [researcher inference · moderate]
• Ownership/terms: Require board information rights, product-roadmap reporting, key-development covenants, operator-recruiting authority and pro-rata rights; no numeric ownership or historical term is supportable. Innovation video, 00:58:58–01:01:36 [researcher inference · provisional]
• Kill criteria: Uncontrolled dependence on an external developer, inability to recruit product leadership, or no executable porting plan for a credible platform transition. Innovation video, 00:58:58–01:01:36 [researcher inference · moderate]
• Next check: Written development rights, roadmap ownership, Noyce or comparable operator response, IBM-platform scenario, porting milestones, and customer retention. Innovation video, 00:58:58–01:01:36 [researcher inference · moderate]
• Reveal — actual decision/outcome: Johnson says he invested early, tried to recruit Bob Noyce, but missed Noyce's warning about Software Arts dependency and did not move onto the IBM PC before Lotus captured the opportunity. His claim that quicker action could have changed the result is counterfactual; vehicle, governance votes, follow-ons, exit and loss remain unknown. Innovation video, 00:58:58–01:01:36 [investor-stated · moderate]

Lab 7 — Genentech access miss

Freeze frame. Byers and Johnson saw the new biotechnology opportunity and wanted access, but the processed accounts say the Swanson/K&P relationship kept the company inside that syndicate. No Johnson thesis memo, offered allocation, or terms are public. CHM Johnson, pp. 23–24 · Perkins, pp. 58–60 [retrospective witness · moderate]

• Invest/pass: Conditional invest if access becomes real and scientific, management, project and financing diligence clear; no analytical pass is evidenced. CHM Johnson, pp. 23–24 [researcher inference · moderate]
• Ownership/terms: Do not model terms without an offered allocation; define minimum governance/information rights before accepting a late syndicate slot. CHM Johnson, pp. 23–24 [researcher inference · provisional]
• Kill criteria: No authentic allocation, inadequate information rights, or inability to diligence science/team independently. CHM Johnson, pp. 23–24 [researcher inference · moderate]
• Next check: Written allocation, data-room access, management/scientific references, cap table and board rights. CHM Johnson, pp. 23–24 [researcher inference · moderate]
• Reveal — actual decision/outcome: Johnson/Byers did not obtain access and invested no capital; Johnson later says Amgen reduced his regret. The record supports an access miss, not a rejected thesis or a lost-return calculation. CHM Johnson, pp. 23–24 [investor-stated · moderate]

Portfolio-level forensics

The 37-family discovery roster is not a portfolio-return sample: 24 are capital-participation leads (including funded VisiCorp, Talkway, and early-AI failures), six are role/list leads without proved target-person capital, two are failure/wrong-selection accounts without proved capital, three are passes, and two are programs. reconciliation · 37-family ledger [researcher inference · strong]

No public fund-level cash flow, gross/net basis, fee/carry method, or vintage-consistent return series was found. Stanford’s SEVF figures are institution-reported program returns, not Johnson-personal returns. Stanford, “Acumen and timing” [researcher inference · strong]

Failure, miss, and anti-portfolio postmortems

The table includes every material failure/miss supported by a reproducible Pitch-specific body. Missing fields are gate failures after the mapped public-route searches, not invitations to infer.

The public record does not support a causal claim that Johnson systematically learned from failures. Missing process change is missing evidence, not proof that no change occurred. failure matrix · CHM Johnson, pp. 19–25 [researcher inference · strong]

Exhausted public-record shortfalls

1. Portfolio denominator: AMC's live biography reports more than 250 investments; HBS counts, the OAC finding aid, direct histories, the official company list, and exact company/name searches together yielded 37 attributable families but no company-level schedule. The 37 is a searched discovery set, not comprehensive coverage. AMC biography · OAC finding aid [researcher inference · strong]
2. D&J/SBIC: the Act, GAO/SBA routes, and Draper/Johnson histories did not expose D&J’s license, debenture, examinations, investment schedule, repayment record, Sutter Hill allocation, or partner accounts. Draper pp. 23–25 remain retrospective testimony. Small Business Investment Act, §§102 and 301–303 · GAO SBIC fact sheet, pp. 9–18 · SBA records-request route · Draper, pp. 23–25 [researcher inference · strong]
3. Amgen: the mapped 1983 AnnualReports URL returns 404 and the issuer archive begins in 2000; searches did not retrieve a public first report/prospectus/cap table/board body. The 2005 auto-captions render a reported 2,685-to-1 figure at 00:55:31–00:55:46, but audio was not independently verified and metric, numerator, denominator, ownership, dilution, distributions, and taxes are undefined; it is retained only as [investor-stated · provisional] and never calculated. Innovation video, 00:55:31–00:55:46 · issuer archive [researcher inference · strong]
4. Tandem: the putative Treybig CHM route returns 404 and exact-title/accession searches did not expose a founder transcript. Perkins pp. 38–43, Byers pp. 8–11, and Johnson p. 18 are the processed boundary; no Treybig words are imputed. attempted CHM route · Perkins, pp. 38–43 · Byers, pp. 8–11 · CHM Johnson, p. 18 [researcher inference · strong]
5. Applied Biosystems: the former CHM URL was John Sanguinetti, not an Applied history. Caruthers’s full participant account supplies formation/product locators but not incorporation, licenses, investor allocations, board rights, ownership or distributions. source-map identity resolution · Caruthers, formation/product paragraphs [researcher inference · strong]
6. VisiCorp/Talkway: the full caption track and Dennis-to-Johnson turn boundary resolve both as distinct target-specific funded failures. The public record still lacks exact dates, legal vehicles, checks, securities, ownership, board votes, complete follow-ons, cash flows, and contemporaneous company records. Innovation video, Johnson turn 00:55:20–01:06:31 [researcher inference · strong]
7. Failures: CHM/Bancroft full-text, company/name variants, official profiles, CHM/OAC/Stanford catalogs, SEC/company routes and contemporaneous-press searches did not identify the early-AI company, Microsoft competitor, online bookseller, or a living-dead company. Early AI stays classified as invested because Johnson says so, but is not promoted to a complete case. CHM Johnson, pp. 19–25 · CHM catalog search · OAC name search [researcher inference · strong]
8. Vehicles and decision rights: Johnson’s 1998 full paper, Bancroft history, SEVF retrospective, RC2 2006 report, current AMC pages, IDEC filing, OAC finding aid and SBA/GAO routes did not expose complete agreements, IC approvals, fees/carry, reserves, follow-on authorization, exit rights, or Johnson cash flows for any row. Johnson Russia paper, printed pp. 7–16 · Bancroft, pp. 35–43 and 78–92 · Stanford SEVF · RC2 reports · IDEC filing, p. 29 · OAC finding aid · SBA records-request route [researcher inference · strong]
9. Archive boundary: the OAC finding aid identifies 78 linear feet/158 boxes, bulk Eastern European records, and company files including Amgen, Tandem, Hypres and Biogen IDEC; item-level bodies were not exposed publicly and some student/employee material is restricted. No archive/person was contacted. OAC finding aid, scope/restrictions [contemporaneous record · strong]
10. Saturation: the final exact-name/company/title sweeps across SEC, CHM, Stanford/HBS/OAC, SBA/GAO, CourtListener and contemporaneous-press routes produced no new complete target-person cash-flow chain. This is a public-access saturation boundary, not a claim about inaccessible private evidence. CHM catalog search · OAC name search · HBS catalog route · SBA records-request route · GAO SBIC fact sheet · CourtListener query [researcher inference · strong]

Coverage and evidence base

The written-corpus denominator is 10 priority routes representing ten candidate works or work families. Six bodies received deep content analysis: four are currently reproducible from live routes (Hoover, the White House compilation, Stanford's 2013 Q&A, and an institutional mirror of the 1998 Stanford/CISAC paper), while the Poland paper and Perkins introduction were processed from full bodies captured in the earlier research pass but are not currently reproducible from their canonical links. Four other routes are partial: a recap rather than the original Entrepreneurial Climate, a limited-preview book chapter/checklist, and two institutionally edited HBS profiles. Treating all six full bodies as live-reproducible would overstate present access. Written-source inventory [researcher inference · strong]

The access classes are deliberately asymmetric. A body captured earlier can support a bounded paraphrase, but current 403 or redirect behavior is recorded so a reader can distinguish evidence used from evidence independently reproducible today. Metadata-only routes are inventoried and analyzed only for what the metadata or excerpt actually says; they are not promoted into unseen prose. coverage ledger · source-map coverage [researcher inference · strong]

Source-by-source analysis

1. Entrepreneurship and Democracy (speech 2004; essay 2005)

Context, authorship, and access. Johnson authored the Hoover essay, adapted from an October 8, 2004 Russian Venture Fair speech; the current Hoover page exposes the full body under stable article headings. This places the argument within his Eastern European institution-building period, but does not establish that the speech and edited essay are word-for-word identical. Closing note

Argument and reasoning chain. The essay starts from a political premise: economic, political, and personal freedoms are mutually reinforcing. Entrepreneurs challenge incumbents, widen consumer choice, create new wealth centers, and therefore diffuse economic power; Johnson then infers that this turbulence supports democracy. Opening; “Entrepreneurs Foster Democracy” [investor-stated · strong] The causal chain is explicit, but its cross-national effectiveness is asserted more than demonstrated. [researcher inference · moderate]

Evidence, examples, and assumptions. Johnson uses US young-company job creation, post-socialist economies, China, Russia, and his own multi-decade observation of technology startups. He also assumes that access to education broadens entrepreneurship and that consumer choice meaningfully disperses power. “Entrepreneurs Foster Democracy” These observations help reveal his worldview; they do not form a controlled comparative test. [researcher inference · moderate]

Decision rules and operating implications. The entrepreneurial-climate section supplies a system checklist: education and research, talent, moderate taxation, venture and bank capital, investment banks/public markets, property and contract law, bankruptcy, and cultural legitimacy. “The Entrepreneurial Climate” [investor-stated · strong] For an investor, the supported rule is to underwrite not only product and founder but also the institutional path from formation through finance and liquidity. Treating “free markets” alone as sufficient would omit half of Johnson’s own list. [researcher inference · moderate]

Biography, decisions, and contradiction. The piece documents Johnson’s attempt to transfer a Valley-derived framework to Russia, while carefully admitting that the post-1990 interval was too short for firm conclusions. Paragraph before “The Entrepreneurial Climate” That caution conflicts productively with the essay’s broad claim that economic freedom ultimately produces political freedom: the mechanism is central, but the timing and boundary conditions remain unresolved. [investor-stated · strong]

2. Promoting Innovation in the USA (2014)

Context, authorship, and access. This signed September 15, 2014 response appears in the official White House Office of Science and Technology Policy compilation, printed pp. 662–665. It is Johnson’s latest full US policy statement in the mapped written corpus. Title/signature, printed pp. 662, 665

Argument and reasoning chain. Johnson distinguishes science (new knowledge whose use may be unclear) from technology and innovation (engineering and market recognition applied to human needs). Research enlarges the option set; entrepreneurs identify the market, complete the engineering, and organize delivery. “Conditions,” printed p. 662 [investor-stated · strong] This is the clearest written expression of his recurring market–technology bridge. Printed p. 662 [researcher inference · moderate]

Evidence, examples, and assumptions. He invokes genetics, electrical materials, atomic structure, logic, university research, angels, venture funds, banks, education, and skilled immigrants. Printed pp. 662–665 The examples support the distinction between knowledge and application, but the paper does not quantitatively test the proposed tax, immigration, or SBIC interventions. [researcher inference · moderate]

Decision rules and operating implications. Two rules transfer directly to investment practice. First, ask separately whether the science is true, the engineering is workable, and a market can be served. Second, align capital duration with a five-to-seven-year company-building interval rather than forcing liquidity to fit the fund calendar. Printed pp. 662–664 [investor-stated · strong] His preferred SBIC design also puts private capital at risk before public money and leaves investment selection private within broad constraints. Printed p. 664 [investor-stated · strong]

Evolution and contradiction. Johnson explicitly adds immigration to the earlier entrepreneurial-climate framework. Printed p. 665 He also favors public support for fundamental research and education while opposing government loan guarantees and insisting on private investment decisions—a bounded-government position, not simple laissez-faire. [investor-stated · strong]

3. Health-Care Entrepreneurship in Poland (2015)

Context, authorship, and access. The nine-page conference paper was mapped to a May 20, 2015 Polish forum. Its body was captured during the earlier research pass, but the StudyLib route returned an HTTP 403 human-verification page on 2026-08-21; the organizer’s event report authenticates the appearance, not every sentence of the mirror. Paper route [contemporaneous record · provisional]

Argument and reasoning chain. The captured paper applies Johnson’s general framework to health care: scientific shifts in genomics, data, and cell-based therapies create possibility; entrepreneurs still need market definition, engineering/product development, teams, capital, regulation, and delivery capacity to turn possibility into patient benefit. Document body pp. 1–9; current route blocked [investor-stated · provisional]

Evidence, assumptions, and operating implications. Its opportunity areas are prospective illustrations, not an audited forecast or portfolio. Document body pp. 1–9; current route blocked [investor-stated · provisional] The useful rule is to map the entire translation chain—science, product, clinical or customer need, regulation, capital, and commercialization—and identify the weakest link before investing; the largest assumption is that Poland can assemble these complements, and the paper body does not establish realized venture returns. Document body pp. 1–9; current route blocked [researcher inference · moderate]

Evolution and contradiction. The 2015 text narrows the 2005 institutional thesis into a sector and country application, while retaining the 2014 separation of science from usable technology. Its narrower domain is evidence of continuity, not proof that the framework selected winners. Because the live body cannot currently be reopened, exact wording, section-level locators, and quotations remain withheld. Poland paper route · 2014 science/technology distinction, printed p. 662 [researcher inference · moderate]

4. Introduction to Tom Perkins oral history (signed 2001; volume 2002)

Context, authorship, and access. Johnson’s signed two-page introduction is an admiring peer assessment embedded in the Bancroft/NVCA oral-history volume. The full introduction was captured/indexed, although the canonical Berkeley file route currently yields the record wrapper in automated retrieval rather than PDF bytes. Printed pp. i–ii

Argument and reasoning chain. Johnson credits Perkins with behaving as an entrepreneur inside venture capital: incubating people and ideas within the firm, organizing outside investors, emphasizing marketing and LP relations, and pairing strong views with honesty and persistence. Printed pp. i–ii [investor-stated · strong] The mechanism is institutional invention plus active company formation, not passive security selection. Printed pp. i–ii [researcher inference · moderate]

Evidence, examples, and attribution. Tandem and Genentech are the main examples. Johnson says Perkins led incubation and external-investor organization; this supports Perkins credit but cannot erase founders Jim Treybig and Bob Swanson, co-investors, or operating teams. Printed p. i [retrospective witness · moderate]

Decision implications and bias. The source suggests that sourcing can be created inside a venture firm and that governance, recruiting, marketing, and syndication are part of the investment product. Printed pp. i–ii [researcher inference · moderate] Yet a ceremonial introduction is structurally favorable. Printed pp. i–ii [researcher inference · moderate] Johnson’s superlative about Perkins’s industry impact is an attributed judgment, not independently verified performance evidence. Printed p. ii [investor-stated · moderate]

5. The Entrepreneurial Climate (1984 recap; original unavailable)

Context and access boundary. Johnson says in 2014 that he wrote the paper in 1984 while teaching at Stanford GSB and that HBS published it. The original text, HBS identifier, and evidence are absent, while a separate profile supplies the unresolved 1982 date. The formerly untitled 1998 route is now identified as the distinct Stanford/CISAC Russia paper; thematic overlap does not establish that it revised The Entrepreneurial Climate. 2014 recap, printed p. 662 · 1998 Russia paper, printed pp. 7–15 [contemporaneous record · insufficient public record]

What can be analyzed. The 2014 recap names seven conditions: low startup regulation; low young-company income/capital-gains taxation; talent and knowledge; venture capital; bank lending; professional services; and social appreciation for entrepreneurs. Printed pp. 662–663 [investor-stated · strong] This establishes the later author’s summary, not the original wording, argument order, citations, or 1984 boundary conditions. Printed pp. 662–663 [researcher inference · strong]

Decision implications and chronology. The recap demonstrates that Johnson had codified ecosystem thinking by his own 1984 account, during his teaching period. It does not prove that the framework caused earlier deals or that later policy additions were present from inception. The correct research action is retrieval, not title-derived synthesis. 2014 recap, printed pp. 662–665 [researcher inference · moderate]

6. The Masters of Private Equity and Venture Capital — chapter and checklist (2010)

Context, authorship, and access. The edited volume contains a Johnson participant chapter, “Fostering Innovation,” and an appendix titled “Checklist for Evaluating Investment Opportunities” under his name. Only contents, limited preview, and exposed appendix material were available; the chapter was not read in full. Book record and contents [contemporaneous record · moderate]

Argument and reasoning visible in the appendix. The checklist proceeds through integrity and guidability, operating experience and complementary skills, market size, technology/product usefulness and defensibility, producibility at workable cost, future financing and valuation plausibility, dilution, and estimated future holding value. Appendix 1, printed pp. 271–272 where exposed [investor-stated · moderate] It is a whole-system underwriting tool rather than a founder-only heuristic. Appendix 1, printed pp. 271–272 where exposed [researcher inference · moderate]

Assumptions and decision implications. The exposed framing acknowledges inadequate information and combines a checklist with judgment. The operational use is to force a written answer for each dimension, construct an independent financing plan, and estimate dilution rather than accepting the founder model at face value. Appendix 1, printed pp. 271–272 where exposed [researcher inference · moderate]

Evolution and contradiction. The checklist formalizes themes expressed more conversationally in the 2013 Stanford Q&A: character, market, differentiation, workability, margins, and active help. Stanford, “What do you look for?” A 2010 codification cannot establish that all items were used ex ante in 1960s–1980s decisions; deal records must perform that test. [researcher inference · moderate]

7. The Application of Venture Capital and the Entrepreneurial Revolution in Russia (1998)

Context, authorship, and identity resolution. Stanford's CISAC published this Johnson-authored paper in August 1998 from his June 5 address to the first annual Russian Venture Capital Association meeting; David Bernstein wrote the foreword, not the argument. The full body is reproducible. Its printed p. 8 contains the sentence excerpted without title in Stanford's 2013 Q&A, resolving that route to this paper rather than an unknown manuscript. Title, foreword, and footnote, printed pp. 3–7 · matching “entrepreneurial revolution” paragraph, printed p. 8 [contemporaneous record · strong]

Argument and reasoning chain. Johnson reconstructs US venture capital as an institution that joined private risk capital, advice, and commitment; places its growth inside a system of SBIC leverage, research universities, talent, taxes, laws, banks, public markets, and social permission; and argues that Russian venture capitalists must help build those complements rather than import capital alone. He explicitly warns that US experience is not automatically portable to Russia. Introduction through “Silicon Valley,” printed pp. 7–10 · Russia application, printed pp. 13–15 [investor-stated · strong]

Evidence, examples, and assumptions. The paper uses Venture Economics figures, Johnson's recollection that 1960s commitments were about $100 million annually or less, 1980's first billion-dollar commitment year, Tandem and Teradyne public-market success, biotechnology and Internet cycles, and Poland's reported venture base. Those observations are useful period evidence but not a controlled test of tax, pension, job, or national-growth causation; figures describe the industry, not Johnson's personal return. Printed pp. 8–13 and figures [investor-stated · moderate]

Decision rules and operating implications. Johnson gives the corpus's clearest written account of early sourcing: visit laboratories, join community groups, speak with lawyers and accountants, and knock on startup doors. Once a lead appears, test people, markets, and working compatibility, then decide whether the investor can supply skill and commitment as well as capital. For a new association, his prescriptions add conduct standards, LP reporting/valuation norms, activity measurement, university training, regional chapters, and cross-border institutional links. Printed pp. 9 and 14–15 [investor-stated · strong]

Evolution, contradictions, and boundary. The paper predates the 2005 democracy essay and already contains most of its institutional complement logic, while treating political transfer more cautiously. It also describes 1990s technology-company valuations as unsupported by long-term earning power, a useful counterweight to an unqualified pro-innovation story. The separate Entrepreneurial Climate title and original 1982/1984 body remain unresolved; thematic overlap is not identity proof. Printed pp. 11–13 [researcher inference · moderate]

8. HBS Alumni Achievement Award profile (2012)

Context, authorship, and access. The HBS award profile is institutionally edited and celebratory, with direct fragments from Johnson. The historic PDF locator is p. 7, but the current PDF URL redirects to the HBS homepage; official HBS profile text remains captured/indexed. This is partial, not live-full, access. HBS award-profile route

Argument, evidence, and examples. The profile is biographical rather than analytical. It frames Johnson’s 1962 move as a deliberate choice to stop remaining solely an employee and uses his father’s post-race handshake lesson to explain an ethic of competition without personal enmity. Profile p. 7 [investor-stated · moderate] Those direct fragments support self-understanding, not a causal claim that athletics produced investment skill. Profile p. 7 [researcher inference · moderate]

Operational implications and assumptions. An analyst may infer a willingness to accept career risk and separate hard competition from relationship destruction, but the source does not connect either trait to a named investment decision. The awards framing selects admirable episodes and contains no misses, cash flows, or countervailing testimony. Profile p. 7 [researcher inference · provisional]

Chronology contradiction. HBS’s 1967 Asset Management date conflicts with Johnson/AMC accounts that place independent investing or firm formation in 1965. The profile is a dated contradiction to preserve, not authority to overwrite entity uncertainty. HBS profile · AMC biography [contemporaneous record · moderate]

9. A founding father of Silicon Valley venture capital guides entrepreneurs (HBS, 2014)

Context, authorship, and access. This short HBS profile/interview contains direct Johnson fragments inside promotional institutional prose. The live route returned HTTP 403 on 2026-08-21; analysis is limited to the official body captured/indexed before that block. HBS article

Argument, reasoning, and examples. Johnson recalls that he and Bill Draper drove through the early Valley, followed promising signs, knocked on company doors, explained venture capital, and offered investment plus help building businesses. Paragraph 2 [investor-stated · moderate] The sequence documents proactive local sourcing and joint attribution; it does not say which visits converted or whether cold outreach generated the best outcomes. Paragraph 2 [researcher inference · moderate]

Decision and biographical implications. The source supports a sourcing rule: do not wait for a polished inbound process when emerging technical activity can be found through direct fieldwork and local networks. Paragraph 2 [researcher inference · moderate] It also situates computer/health-care investing and teaching in Johnson’s later identity, but its more-than-200 company count is not an audited denominator. Opening/closing paragraphs [contemporaneous record · provisional]

Contradictions. The profile again uses 1967 for Asset Management and a smaller portfolio claim than AMC’s more-than-250. Dates, entities, follow-ons, and inclusion rules are not supplied, so neither number can anchor coverage or hit rates. HBS profile · AMC biography [researcher inference · moderate]

10. Venture Capital Pioneer Illuminates the Silicon Valley Ecosystem (Stanford Q&A, 2013)

Context, authorship, and access. Stanford’s full written Q&A presents Johnson’s direct answers under question headings, surrounded by institutional biography and promotional statistics. The body is currently live and reproducible. Full Q&A

Argument and selection chain. Johnson’s sequence begins with integrity, honesty, openness, and decency; then market vision; then differentiation, product workability, competitive insulation, and margins. He contrasts this people-first ordering with Don Valentine’s market-first ordering while acknowledging that both require people, market, and a differentiated product. “What do you look for?” [investor-stated · strong] The disagreement is about ordering and replaceability, not whether markets matter. “What do you look for?” [researcher inference · strong]

Evidence, examples, and biography. Johnson names Amgen as his proudest investment, credits himself alongside Bill Bowes, and emphasizes George Rathmann and patient benefit. “Of which investment are you most proud?” [investor-stated · moderate] This is retrospective credit allocation; it does not establish legal founding, sourcing, check size, ownership, or returns. Same Stanford answer [researcher inference · strong]

Operating rules. For founders: competitive drive, intuitive market understanding, technical depth, integrity, and zeal. For VCs: retain the ability to make small initial commitments, avoid impatience, and help with customer access, strategy, tactics, plans, and mundane execution. Entrepreneur-characteristics, VC-mistakes, and founder-VC questions [investor-stated · strong]

Failure, forecast, and limits. Johnson says zeros often followed failure to see a founder’s inability to work with others or build commercial relationships, and he identifies running out of money through optimistic growth plans as a common founder error. Worst-investment and founder-mistakes questions [investor-stated · moderate] His forecast of mobile/cloud, genomics, and cell therapies was ex ante in 2013, but it is broad enough that later technological progress alone would not prove investment selection skill. [researcher inference · moderate]

Synthesis: the written doctrine

Across the corpus, five recurring beliefs form a coherent doctrine, but repetition establishes consistency rather than efficacy. White House response, printed pp. 662–665 · Stanford Q&A [researcher inference · moderate]

Portfolio-function implications

• Sourcing: search where technical knowledge, operating talent, and unmet markets intersect; early D&J door-knocking is a concrete joint-sourcing example. HBS 2014, paragraph 2 [researcher inference · moderate]
• Picking: score character, market vision, differentiation, workability, producibility, margins, financing needs, and dilution independently. Stanford selection question Book appendix [investor-stated · moderate]
• Pricing and ownership: the exposed checklist requires a five-year holding-value estimate after dilution, but the corpus provides no general ownership target or valuation ceiling. Book appendix [investor-stated · moderate]
• Portfolio construction: Johnson criticizes large firms that cannot do small deals and fund structures that force time pressure; the writings do not disclose concentration, reserves, or loss budgets. Stanford VC-mistakes question RFI p. 663 [investor-stated · strong]
• Governance: active help with customers, strategy, tactics, plans, recruiting/incubation, and syndication is supported; contractual control, board escalation, and removal thresholds are not. Stanford founder-VC question Perkins intro, p. i [investor-stated · moderate]
• Exits: public markets appear as an ecosystem liquidity mechanism and the checklist estimates future holding value, but no general sell discipline is stated. Hoover climate section Book appendix [investor-stated · moderate]

Intellectual chronology, evolution, and contradictions

1. By 1984, according to Johnson’s later self-date: he had codified seven entrepreneurial-climate conditions while teaching at Stanford. Only the 2014 recap survives here, so its original argument and whether an earlier 1982 version existed remain unknown. RFI p. 662 [contemporaneous record · moderate]
2. 1998: the full Stanford/CISAC Russia paper describes the US history, Johnson's early search/process, 1990s valuation excess, ecosystem complements, and a cautious Russian transfer program. Its printed p. 8 exactly supplies Stanford's later excerpt, resolving that identity; it does not resolve the separate 1982/1984 Entrepreneurial Climate body. Russia paper, printed pp. 7–15 [contemporaneous record · strong]
3. 2001–2005: the Perkins introduction articulates active firm-building through a peer; the Hoover essay enlarges the lens to law, capital, education, social permission, and democracy. Perkins intro Hoover essay [researcher inference · moderate]
4. 2010–2013: the checklist and Stanford Q&A operationalize judgment at company level—people, market, product, finance, dilution, patient help—and preserve an explicit disagreement with Valentine on ordering. Book appendix Stanford selection question [investor-stated · moderate]
5. 2014–2015: Johnson updates the ecosystem model with skilled immigration, fund-duration concerns, and a US policy program, then applies related reasoning to Polish health care. RFI pp. 663–665 Poland route [investor-stated · moderate]

Three tensions should remain visible. First, people-first versus market-first is unresolved, though Johnson states the disagreement precisely. Second, patience versus stopping is incomplete: he criticizes impatience but gives no written kill criteria. Third, bounded government versus economic freedom is not a contradiction once decomposed: Johnson wants public research, education, immigration reform, and legal infrastructure, while keeping private investors first-loss and in control of selection. Hoover climate/conclusion RFI pp. 663–665 [researcher inference · moderate]

Modern VC translation (as of 2026-08-20)

Each transfer needs its own present-day test; these are analyst-created applications, not Johnson deal accounts. written-doctrine synthesis [researcher inference · moderate]

• Science, engineering, and market — [researcher inference · moderate] [analyst-created hypothetical]. For a university spinout with promising lab results but no validated buyer, write separate scientific-validity, engineering/manufacturing, and customer/regulatory memos; fund only the cheapest independently disconfirming milestone. White House response, printed pp. 662–665
• People and paper protections — [researcher inference · moderate] [analyst-created hypothetical]. For a first-time founder whose references are mostly friendly investors, require structured former-colleague references, a bad-news simulation, and observed treatment of dissent before treating integrity as underwritten. Stanford founder-characteristics answer
• Check size and patience — [researcher inference · moderate] [analyst-created hypothetical]. For a capital-intensive company with a two-year technical proof point, model whether the proposed first check, reserve budget, and fund life survive both a six-month slip and one extra round; pass if vehicle pressure would force premature scale or abandonment. White House response, printed p. 663 · Masters appendix
• Ecosystem complements — [researcher inference · moderate] [analyst-created hypothetical]. For a regional-cluster thesis, name the missing customer, talent, legal, research, financing, and liquidity complements and invest only if the company or syndicate can close the binding gap rather than merely copy a Silicon Valley label. Hoover climate section · Russia paper, printed pp. 13–15
• Active help and founder agency — [researcher inference · moderate] [analyst-created hypothetical]. Before accepting a board seat, agree with the founder on one investor-owned bottleneck, the decisions that remain management's, a measurable result, and a date to stop the intervention. Stanford founder-VC answer · Perkins introduction, p. i
• Financing and dilution — [researcher inference · moderate] [analyst-created hypothetical]. For a company proposing an optimistic single-round plan, run base, slow, down-round, and extra-round cases; compare surviving ownership and control with the evidence still required, and do not let spreadsheet precision substitute for customer proof. Masters appendix

Best writings, study order, and unresolved retrievals

Read these five first

1. Promoting Innovation in the USA (2014). Best compact statement of the science–technology–market distinction, the seven-part ecosystem, fund-duration tension, and Johnson’s policy boundary. Printed pp. 662–665 [researcher inference · strong]
2. The Application of Venture Capital and the Entrepreneurial Revolution in Russia (1998). Best contemporaneous written bridge from Johnson's early sourcing and hybrid-capital practice to industry history, valuation caution, and institution-building. Printed pp. 7–15 [researcher inference · strong]
3. Stanford GSB Q&A (2013). Best operational direct voice on character versus market, founder traits, margins, patience, small deals, active help, failures, and forecasts. Question-heading index [researcher inference · strong]
4. Entrepreneurship and Democracy (2005). Best full account of the institutional worldview and its largest empirical vulnerability. Article headings [researcher inference · strong]
5. The Masters checklist (2010, partial). Best structured investment-screen route, but rank it below the full bodies until lawful full-text access resolves authorship/editorial boundaries. Book record and exposed appendix [researcher inference · moderate]

Unresolved written record

• The Entrepreneurial Climate: obtain the original HBS publication/case number and body and reconcile Johnson's 1984 self-date with the 1982 reference. The distinct 1998 Stanford/CISAC Russia paper is now identified; thematic overlap does not establish that it is a revision of the missing work. Until retrieval, only the 2014 recap is content evidence for The Entrepreneurial Climate. RFI p. 662 · 1998 paper, printed pp. 7–15
• Masters chapter/checklist: obtain lawful full text and separate Johnson’s authored material from editorial reconstruction. Book record
• Poland paper: obtain an organizer/original-file copy; the StudyLib mirror currently returns 403, so exact text and page locators are not independently reproducible. Current route
• HBS 2012 and 2014 profiles: recover stable official or archived copies; the older PDF redirects and the 2014 page returns 403. Their direct fragments remain usable as bounded institutional evidence, not as full current bodies. [researcher inference · moderate]
• Decision evidence: no public investment memos, board materials, partnership letters, complete portfolio schedules, reserve decisions, ownership targets, or written stop rules were found in this corpus. These omissions prevent the writings alone from proving that the stated doctrine drove results. archive and unresolved-record routes [researcher inference · insufficient public record]

Coverage and evidence boundary

The source map contains 17 priority spoken-source rows collapsed into 14 appearance families. Seven families have full public bodies processed here: four official, speaker-identifiable transcripts—the 1997 Boole & Babbage history, 2008 Bancroft history, 2008 Dennis–Johnson Venture Voices interview, and 2019 CHM history—and three official CHM videos with complete machine-generated en-orig caption tracks—Pioneers of Venture Capital, Innovation Is Nothing New, and Blazing a Trail. In the videos, only Johnson turns bounded by direct address, explicit handoff, or unmistakable self-identification are attributed to him; machine captions support paraphrase and timestamps, not exact quotation. Seven families remain partial, restricted, edited, or metadata-only: Something Ventured, GTS 2009, the 2010 RT interview, the 2010 Moscow lecture, BioForum 2015, MVS25, and SIEPR 2010. Each family has a canonical note. [researcher inference · strong] source inventory · video-control notes

The richest evidence is retrospective. It reliably establishes what Johnson said and how he narrated his process at the interview date; it does not by itself prove what he knew at the investment date, his ownership, or cash-flow contribution. Repetition across interviews establishes narrative consistency, not independent corroboration or efficacy. [researcher inference · strong] Bancroft rights/custody, catalog and transcript · CHM oral history, PDF pp. 1–2

Source-by-source guide

1997 — Boole & Babbage oral history

This is the best deal-specific spoken record. It traces founders Ken Kolence and Dave Katch, John Bryan's referral, Johnson's participation in formation and financing, a remembered roughly $250,000 initial financing, the customer thesis, IBM conflict, maintenance economics, management changes, platform obsolescence, and later renewal. [investor-stated · moderate] CHM transcript, printed pp. 4–10

The investment rule is narrower than “back software”: productize an expensive recurring customer problem, provide service that a free orphan product cannot, and plan for product two. The case also warns that long governance involvement may be skill, sunk-cost persistence, or both; the public record lacks the ownership and cash-flow chain to choose cleanly. [researcher inference · moderate] CHM transcript, printed pp. 7–11

2002 — Pioneers of Venture Capital

The full CHM video yields three speaker-controlled Johnson turns. After a direct moderator handoff, Johnson links his early judgment apprenticeship to production work at Inland Steel and advises students to spend time making or designing something before entering finance. That is his career heuristic, not proof that operator experience caused investment returns. [investor-stated · moderate] video, 00:29:50–00:33:20

The deal discussion separates credit: Perkins says Tandem was incubated inside Kleiner Perkins and calls Johnson the validating outside investor; Johnson credits Brook Byers with persuading him. Johnson contrasts the thirty-year Boole & Babbage hold, enabled in his account by private capital, with Tandem's more conventional time to success. [researcher inference · strong] video, 00:51:59–00:54:19 Asked about the dot-com crash, Johnson argues that a company still needs a path to sales, earnings, and growth; the diagnosis is contemporaneous to 2002 but not a deal-by-deal record of his own bubble exposure. [investor-stated · moderate] video, 01:18:15–01:20:52

2005 — Innovation Is Nothing New

This full video materially improves the failure record. Reid Dennis directly asks Johnson about a bad deal at 00:55:20; Johnson remains the speaker through the Amgen, VisiCorp, Boole & Babbage, and Talkway sequence until Dennis responds at 01:06:31. The caption text is machine-generated, so the note retains paraphrase rather than exact quotes. [researcher inference · strong] video, 00:55:20–01:06:31 · source note

Within that turn, machine captions render a reported 2,685-to-1 Amgen figure at 00:55:31–00:55:46. The audio was not independently verified, and the metric, numerator, denominator, ownership, dilution, distributions, and taxes are undefined, so the figure is retained only as a caption-mediated report and never as a verified personal return. [investor-stated · provisional] video, 00:55:31–00:55:46

VisiCorp is the stronger postmortem. Johnson says VisiCorp depended on Software Arts for VisiCalc product development; Bob Noyce warned him about the dependency; and failure to move onto the IBM PC before Lotus destroyed the opportunity. Johnson admits he missed the warning, but his claim that faster action could have challenged Lotus is counterfactual and the source supplies no board or product record. [investor-stated · moderate] video, 00:58:58–01:01:36

Talkway is the more honest but less complete diagnosis. Johnson says a failed advertising-services model led to a video-email pivot and more investment because the engineers and use case remained compelling; customers still did not adopt. He says he did not fully understand why, while accepting a major market misjudgment. This supports a demand error and possible persistence problem, not a precise causal verdict. [researcher inference · moderate] video, 01:04:29–01:06:31

2008 — Bancroft oral history

Across three sessions, Johnson gives the fullest biography and the broadest vehicle/deal record. He distinguishes D&J, post-sale personal investing, a family SBIC, later partnerships, and outside/family structures. [investor-stated · moderate] Bancroft, printed pp. 20–42 and 64–79 He separately narrates Tandem and Amgen, teaching and policy, and Eastern Europe, exposing ambiguities a celebratory profile would erase. [researcher inference · moderate] Bancroft, Tandem/Amgen at printed pp. 39–63, Eastern Europe at pp. 78–92, and teaching/policy at pp. 94–100

Two moments matter for decision practice. First, Amgen began with a broad scientific possibility set; Johnson says he did not know it was a winner, only that the area was large, and the company later concentrated on EPO. [investor-stated · moderate] Bancroft, 03-00:52:31–04-00:11:24; printed pp. 55–60 Second, he asks whether he was too patient with management at Boole & Babbage. [investor-stated · moderate] Bancroft, 04-00:19:01; printed pp. 62–63

2008 — Reid Dennis and Pitch Johnson, Venture Voices

The joint transcript gives Johnson's most compact operating formulation: back people able to turn technology into products for large markets; contribute money, time, and emotional commitment; replace management when required; stop adding capital when the return case no longer works. [investor-stated · moderate] Venture Voices, 02:21:20–02:27:40; tape 2 pp. 16–20

Speaker control matters. Dennis—not Johnson—delivers the extended people-judgment discussion about a decision frame and much of the excess-capital critique. Johnson separately contrasts early personal/family capital with institution-backed funds and calls Perkins the first Tandem VC and himself second. [researcher inference · strong] Venture Voices, 02:06:29–02:08:00 and 02:19:20–02:24:29

2009 — Something Ventured interview route

Stanford's collection metadata states that raw digital interview transcripts exist, while public item-level access is restricted. The distributor study guide describes the edited documentary but cannot substitute for Johnson's raw interview. No raw-interview content is inferred from the film. [researcher inference · strong] Stanford collection · official guide

2009 — Global Technology Symposium highlights

The contemporaneous report verifies a panel with Johnson, Draper, Dennis, and Winblad. Because the article summarizes the group without a full speaker transcript, it supplies appearance context only. [researcher inference · strong] TechCrunch

2010 — SIEPR personalized-medicine panel

The official program records Johnson as moderator alongside Brook Byers and Elizabeth Holmes. A moderator credit does not establish endorsement, diligence, an investment relationship, or later knowledge; no body was found. [researcher inference · strong] SIEPR program

2010 — Moscow lecture

The RIA report verifies a public lecture on entrepreneurship, markets, government, and ethics but lacks an original recording or speaker-checked transcript. It supports appearance and reported-theme context, not exact Johnson wording or regional-fund economics. [contemporaneous record · moderate] RIA report, article paragraphs

2010 — Russia interview

The RT interview supplies edited attributed excerpts on domestic markets, export sequencing, capital, corruption, and institutional transfer. It shows Johnson applying the market framework outside the United States, but the edit boundary and venue framing remain material and the source documents no regional-fund economics. [investor-stated · moderate] RT interview, article paragraphs

2015 — BioForum, Wroclaw

The event report authenticates Johnson's appearance and supports the provenance of the separately analyzed health-care paper. It does not prove the written paper was delivered line-for-line. [researcher inference · strong] BioForum report

2019 — Blazing a Trail

The official full video adds three bounded Johnson turns. Johnson credits Bowes with the Amgen opportunity and invitation and Rathmann with building the company, but the short account omits Salser and Alstetter and therefore cannot settle formation credit. [researcher inference · strong] video, 00:07:08–00:10:00 · Bancroft, printed pp. 47–55

Johnson later describes family/individual capital and says he and his wife were the limited partners behind a separately managed firm; no legal-entity or economics schedule appears. [investor-stated · provisional] video, 00:53:48–00:54:12 He then uses an unidentified early AI investment to argue that technology must meet an existing or credibly created market. The next speaker explicitly marks the turn boundary by agreeing with Johnson. [investor-stated · moderate] video, 00:56:38–00:57:56

2019 — CHM oral history

The latest full interview is valuable less because it is new than because it makes attribution explicit. Johnson credits Bowes for the “applied” deal invitations, Perkins and Treybig for Tandem formation, John Bryan for referring Kolence and Katch, and colleagues for co-investing. [investor-stated · moderate] CHM, Tandem at PDF p. 18, Bowes at pp. 19–20, and Boole at pp. 21–23 The same interview preserves a negative attribution result: Johnson says he mixed up the order and direction of the Coherent/Rockefeller connection, so it cannot establish who sourced whom. [investor-stated · moderate] CHM, PDF pp. 15–17

It also gives the clearest adverse inventory. Intel was allocation-constrained. CHM, PDF pp. 14–16 Genentech was sought but unavailable; Apple was not pursued eagerly enough because Johnson underestimated household use, although he resists a simple characterization as having missed it; and a deliberately built Microsoft competitor lost. CHM, Genentech/Apple/Microsoft at PDF pp. 23–25 An unidentified AI investment was at least fifteen years before the interview and, in Johnson's account, roughly twenty years early; identity, vehicle, and terms remain unknown. CHM, PDF pp. 20–21 These are different decision errors or constraints, not one category of miss. [researcher inference · moderate]

2025 — MVS25 fragment

The host-posted clip contrasts technical expertise with judging grit and determination. It is directionally consistent with the older people-plus-market rule, but the selected fragment and missing full program make it weak evidence of late evolution. [investor-stated · provisional] host post

Recurring frameworks across the spoken corpus

1. People convert technology into served markets. The 2008 joint interview and 2019 Blazing panel repeat the conjunction: technology is an input; capable people must use it to serve an existing or credibly created market. [investor-stated · moderate] Venture Voices, 02:21:20–02:23:00 · Blazing video, 00:56:38–00:57:56
2. Learning expands the investable map. Johnson links continuing computer and molecular-biology coursework to later opportunity recognition. That is self-attributed causation and should be tested against contemporaneous course/deal files. [investor-stated · moderate] Bancroft, printed pp. 28–31, 45–47
3. Help is a controlled intervention, not shadow management. Weekly contact, board committees, recruiting, and management change appear across Boole, Tandem, and Amgen. The counter-risk is overinvolvement or patient capital becoming sunk-cost persistence. [researcher inference · moderate] Boole transcript, printed pp. 5–10 · Bancroft, Tandem at printed pp. 39–45 and Amgen at pp. 50–63
4. Capital source changes behavior. Johnson contrasts personal/family capital with institution-backed money; the 2002 and 2019 panels make the duration and family-LP claims explicit. The causal claim—patient capital enabled long-duration intervention—remains a researcher inference because comparative vehicle cash flows are absent. [researcher inference · moderate] Pioneers video, 00:53:12–00:54:19 · Blazing video, 00:53:48–00:54:12
5. Failure modes differ. VisiCorp combines an ignored dependency warning with platform-transition delay; Talkway combines capable people, a financed pivot, absent adoption, and an unresolved causal diagnosis. These join exclusion (Genentech), allocation (Intel), imagination/urgency (Apple), premature timing (the unidentified AI company), competitive execution (the Microsoft rival), and management lag (Boole). Each failure requires a different process correction; collapsing them into “bad picks” loses information. [researcher inference · moderate] Innovation video, 00:58:58–01:06:31 · CHM, PDF pp. 20–25

Evolution, contradictions, and decision evidence

The spoken record is stable in its people-plus-market core but grows more explicit about failure and attribution. The 1997 Boole interview emphasizes product economics, platform exposure, and management succession. In 2002 Johnson adds operator apprenticeship, patient-capital duration, and earnings discipline after the dot-com crash. In 2005 he gives unusually specific VisiCorp and Talkway postmortems. The 2008 histories foreground vehicles, governance, and emotional labor; the 2019 appearances foreground team credit, early-AI timing, and anti-portfolio memory. [researcher inference · moderate] Boole transcript, printed pp. 4–11 · Pioneers video, 00:29:50–00:33:20 and 01:18:15–01:20:52 · Innovation video, 00:58:58–01:06:31 · CHM, PDF pp. 18–25

Four tensions should remain unresolved:

• Patience versus stop discipline. Johnson advocates walking away from nonworking deals, questions whether he waited too long to change Boole management, and says Talkway received more capital after its first model failed. The record does not show an ex-ante stop rule. [researcher inference · moderate] Venture Voices, 02:26:00–02:27:40 · Bancroft, 04-00:19:01 · Innovation video, 01:04:29–01:06:31
• People first versus market first. The synthesis is conjunctive, not a choice: people must serve a large market. Talkway is direct counterevidence to treating sincere, capable builders as sufficient. [researcher inference · strong] Venture Voices, 02:22:43 · Innovation video, 01:04:29–01:06:31
• Specific warning versus story after the fact. VisiCorp contains a remembered Noyce warning and a platform mechanism; Talkway ends with admitted uncertainty. Both are retrospective and lack contemporaneous board evidence, so VisiCorp's cleaner lesson may partly reflect narratability. [researcher inference · moderate] Innovation video, 00:58:58–01:06:31
• Company builder versus one-outlier economics. The spoken record celebrates hands-on work across many companies while Johnson says Amgen alone could have carried the record. Without a cash-flow denominator, neither narrative wins. [researcher inference · moderate] CHM, PDF pp. 19–21

Listen to these five first

1. Bancroft oral history — best complete life, vehicle, Amgen, teaching, policy, and Eastern Europe record; read with the rights and retrospective-memory caveat. note
2. CHM 2019 oral history — best concise cross-check, explicit attribution, and differentiated miss inventory. note
3. Innovation Is Nothing New — best single spoken failure sequence for VisiCorp, Talkway, portfolio skew, and the limits of retrospective causal clarity; use paraphrases, not machine-caption quotations. note
4. Boole & Babbage oral history — closest deal-specific retrospective account of the product thesis, platform risk, management succession, and long-hold case. note
5. Venture Voices — best compact people/market/board/stop framework, with strong speaker labels. note

Retrieval priorities and study boundary

The next valuable retrievals are the raw Something Ventured interview; the full MVS25 program; the SIEPR panel; and the Moscow lecture. An official speaker-labeled transcript or manual audio verification for the three CHM captioned videos would permit exact quotations and improve unclear-name control, but the full bodies and bounded Johnson turns have now been processed. [researcher inference · strong] Something Ventured route · SIEPR program

Use this spoken record in sequence with the written corpus, deal ledger and decision labs, operating system, and mental models. A reader should use the transcripts to hear Johnson's own distinctions, then use company/filing evidence to test attribution and economics.

1. Back people who can turn technology into a large served market

• Rule — [researcher inference · moderate] Fund a capable, trustworthy team only when it can explain whose problem the technology solves, why the market can become large, and how the company—not the technology alone—will reach that market. The combined decision rule is a synthesis of Johnson’s separate people, integrity, and market statements. CHM oral history, PDF pp. 19–21 · Dennis–Johnson Venture Voices, 02:21:32–02:24:29
• Evidence status — [researcher inference · moderate] Johnson states the people and market components directly in two late oral histories; the combined rule is reconstructed. Application evidence survives for several celebrated cases and the bounded Talkway failure, but the public record is retrospective and the 37-family discovery roster is selected by source survival. CHM, PDF pp. 19–21 · CHM Innovation Is Nothing New video, 01:04:29–01:06:31
• Provenance and origin — [retrospective witness · moderate] Johnson remembered Georges Doriot teaching that technology has value only if somebody wants it; the same interview distinguishes useful research from market-serving technology. CHM oral history, PDF pp. 9 and 20–21
• Mechanism and reasoning — [investor-stated · moderate] Technical novelty is an input, not demand. The entrepreneur must identify a real use, translate the technology into a product and organization, recruit followers, and reach a market large enough to repay venture risk. Dennis–Johnson Venture Voices, 02:21:32–02:24:58
• Historical and vehicle context — [researcher inference · moderate] The rule emerged when small syndicates financed hardware, packaged software, and biotechnology before mature category playbooks existed; sparse capital made market selection and hands-on company formation unusually important. This is a contextual reconstruction, not a Johnson-authored causal proof. Boole & Babbage oral history, printed pp. 7–11 · Byers oral history, printed pp. 9–10/PDF pp. 18–19
• When to use — [researcher inference · moderate] Analyst-created application checklist: Use for science- or engineering-led seed decisions where product utility and distribution remain uncertain. Do not use “large market” as a top-down category slogan when the team cannot name a buyer, urgent use, adoption path, or value-capture mechanism. CHM oral history, PDF pp. 19–21
• How to apply it — [researcher inference · moderate] Analyst-created application checklist: (1) Name the user and paid problem; (2) separate research, technology, product, company, and market risk; (3) test whether the founder can recruit and sell; (4) size the reachable market from a bottom-up adoption path; (5) kill the deal if willingness-to-pay, route to market, or leadership evidence remains purely asserted; (6) set the next check at a paid-use or repeat-use milestone. CHM oral history, PDF pp. 19–21 · Talkway failure, 01:04:29–01:06:31
• Examples
  • Tandem — researcher-applied illustration. [researcher inference · moderate] Johnson later praised Jim Treybig for spotting a market and meeting projections, while Brook Byers independently says he performed the NonStop Computing diligence and Perkins controlled admission to the syndicate. The case illustrates people-plus-market fit, not a Johnson-only pick. Dennis–Johnson, 02:23:57–02:24:29 · Byers, printed pp. 9–10/PDF pp. 18–19
  • Boole & Babbage — researcher-applied illustration. [researcher inference · moderate] Kolence and Katch saw a market for packaged software that improved IBM systems; Johnson and other investors helped organize the company and governance. The surviving record connects the use case to the company, but not to a formal pre-investment memo. Johnson CHM, PDF pp. 21–23 · Boole & Babbage oral history, printed pp. 7–8
  • Talkway — demand-absence failure/boundary. [retrospective witness · moderate] Johnson says his investing group helped start Talkway, funded a pivot from unsuccessful advertising services to video email, and found that demand still did not materialize; he adds that he did not fully understand the failure. The episode supports a demand-test boundary, not a complete causal diagnosis. Speaker-checked machine captions support this paraphrase, not quotation. CHM Innovation Is Nothing New video, 01:04:29–01:06:31
• Evidence — [investor-stated · moderate] Two direct interview families repeat the people-plus-large-market rule. [researcher inference · moderate] Independent Byers and company-specific Boole testimony support parts of its application; the speaker-checked Talkway account adds a bounded demand-failure counterexample, while the missing complete portfolio prevents an efficacy test. Dennis–Johnson, 02:21:32–02:24:29 · Byers, printed pp. 9–10/PDF pp. 18–19 · Boole & Babbage oral history, printed pp. 7–10 · CHM Innovation Is Nothing New video, 01:04:29–01:06:31
• Limits, failure modes, and exception ledger — [researcher inference · insufficient public record] Large-market claims can conceal poor entry economics, slow adoption, capital intensity, or no distribution. Apple shows Johnson could underestimate a market; Talkway shows a funded product pivot can still meet absent demand without yielding a complete causal diagnosis. Attempted routes and results: CHM and Bancroft oral histories yielded retrospective accounts; speaker-checking the full CHM video yielded Johnson's uninterrupted Talkway answer; the Stanford/Something Ventured raw-interview route remains restricted; no Talkway investment memorandum, board packet, terms, or customer study surfaced. CHM, PDF pp. 23–25 · CHM Innovation Is Nothing New video, 01:04:29–01:06:31 · Something Ventured archival record
• Evolution and contradictions — [investor-stated · moderate] Johnson retained the people/market rule late in life, yet also acknowledged that he lacked imagination about Apple's possible use and did not pursue it eagerly enough; he explicitly resisted a simple characterization as having missed it. CHM, PDF pp. 23–24
• Modern VC translation (as of 2026-08-21) — [researcher inference · moderate] [analyst-created hypothetical] Analyst-created current transfer hypothesis: prompted by Talkway's reported demand failure, require observable buyer behavior, retention, or a credible procurement path before treating a pivot or category narrative as company-level demand. This is a decision-test hypothesis, not a claim about all current markets. Hypothetical: between two AI workflow startups with identical model quality, prefer the one with recurring paid use that persists after an incumbent bundles a similar feature. CHM Innovation Is Nothing New video, 01:04:29–01:06:31

2. Learn enough science to test the business, not to impersonate the scientist

• Rule — [researcher inference · provisional] Build enough technical fluency to identify the decisive scientific risk and ask informed questions, then require market, organizational, and financing evidence before investing. Bancroft oral history, printed pp. 41–46
• Evidence status — [researcher inference · provisional] Johnson directly describes continuing technical study and delight in technology, but the combined rule is reconstructed. Three illustrations survive, while independent evidence corroborates the Boole product and market rather than Johnson’s pre-investment technical-diligence process. Kolence oral history, printed pp. 18–21
• Provenance and origin — [retrospective witness · moderate] Johnson linked his engineering background and later Stanford courses to understanding new technical fields, while warning that venture capitalists can merely sound knowledgeable. Bancroft oral history, printed pp. 41–42
• Mechanism and reasoning — [researcher inference · provisional] Technical fluency may improve question quality and expert selection; it does not validate demand, leadership, regulation, capital needs, or investment economics. The safeguard is to translate each technical assertion into a falsifiable milestone and an economic consequence. Bancroft, printed pp. 41–46 · Kolence, printed pp. 18–21
• Historical and vehicle context — [researcher inference · moderate] Early laser, computing, and recombinant-DNA markets lacked today's specialist funds and data rooms. Johnson's family-capital practice could pair generalist board work with outside scientists and operators, but exact diligence budgets and decision rights remain unavailable. Bancroft, printed pp. 41–46
• When to use — [researcher inference · moderate] Analyst-created application checklist: Use when feasibility, manufacturability, clinical translation, or system reliability is a gating risk. Do not use personal technical enthusiasm as a substitute for independent expert review or customer evidence. Bancroft oral history, printed pp. 41–46
• How to apply it — [researcher inference · moderate] Analyst-created application checklist: (1) State the technical claim in testable form; (2) identify the qualified expert and conflict; (3) specify replication, prototype, reliability, regulatory, or yield evidence; (4) translate success into customer utility; (5) map capital and time to the milestone; (6) kill if the team cannot disclose disconfirming evidence or the investor cannot obtain competent review. Bancroft oral history, printed pp. 41–46 · Kolence oral history, printed pp. 18–21
• Examples
  • Amgen — researcher-applied illustration. [researcher inference · moderate] Johnson's account says Bowes and Sam Alstetter formed Applied Molecular Genetics, Winston Salser became the forming scientist, Bowes invited Johnson, and Bowes and Johnson recruited George Rathmann after reference work. Johnson reports a small family-partnership investment; legal identity, ownership, and economics remain unverified. Bancroft, printed pp. 42–46
  • Boole & Babbage — researcher-applied illustration. [researcher inference · moderate] The company tied software performance to a customer’s ability to postpone mainframe upgrades. Johnson’s account describes the economics; founder Ken Kolence independently describes the product, marketing test, customer budget constraint, and competitive response. Neither source preserves a Johnson pre-investment memo. Boole & Babbage oral history, printed pp. 7–8 · Kolence oral history, printed pp. 18–21
  • VisiCorp — product-control and platform-transition failure/boundary. [retrospective witness · moderate] Johnson recalls early participation, Noyce's warning about relying on a product the company did not control, and failure to adapt for the IBM PC; he offers Lotus as a hindsight counterfactual. The episode supports expert-listening and dependency tests, not proof that one decision alone caused the outcome. Speaker-checked machine captions support this paraphrase, not quotation. CHM Innovation Is Nothing New video, 00:58:58–01:01:36
• Evidence — [retrospective witness · provisional] Bancroft documents Johnson's technical-study rationale and the Amgen recruitment sequence. [researcher inference · provisional] Kolence independently corroborates Boole's product and market, while the speaker-checked VisiCorp account supplies a direct retrospective counterexample about product control and platform transition; neither reconstructs Johnson's full pre-investment technical-diligence process. Bancroft, printed pp. 41–46 · Kolence, printed pp. 18–21 · CHM Innovation Is Nothing New video, 00:58:58–01:01:36
• Limits, failure modes, and exception ledger — [researcher inference · insufficient public record] Technical fluency can produce confidence theater and domain overreach. Attempted routes and results: the full Bancroft and CHM Johnson transcripts yielded retrospective process; Kolence independently yielded Boole product/market facts but not Johnson's diligence; speaker-checking the full CHM video yielded the VisiCorp product-control warning and IBM-PC transition account; the Amgen annual-report archive yielded company records but not private pre-investment work. No Johnson memo, VisiCorp board record, expert report, or contemporaneous IC record surfaced. Amgen annual-report archive · CHM Innovation Is Nothing New video, 00:58:58–01:01:36
• Evolution and contradictions — [retrospective witness · moderate] Johnson continued studying science, but by 2019 said he no longer had the patience for detailed deal homework and referred opportunities to a specialist team. That is an evolution toward role specialization, not proof that earlier diligence was complete. CHM, PDF p. 17
• Modern VC translation (as of 2026-08-21) — [researcher inference · moderate] [analyst-created hypothetical] Analyst-created current transfer hypothesis: prompted by VisiCorp's reported product-control warning and platform-transition miss, pair technical fluency with dependency mapping, conflict-checked outside review, and falsifiable evidence appropriate to the product. This is a diligence hypothesis, not a claim that every current technical market shares the same dependencies. Hypothetical: for a gene-editing seed deal, require independent replication and a regulator-informed development path before a tranche tied to human-study preparation. CHM Innovation Is Nothing New video, 00:58:58–01:01:36

3. Underwrite integrity, followership, and the next leadership transition

• Rule — [researcher inference · moderate] Back leaders with demonstrated integrity and the ability to attract followers, but decide before investing how the board will recognize and handle a founder-to-executive transition. Integrity and followership are Johnson-stated; the precommitment is researcher synthesis. CHM, PDF pp. 19–20 and 22–23 · Dennis–Johnson, 02:25:31–02:27:01
• Evidence status — [researcher inference · moderate] Integrity and followership are direct criteria, while the transition rule is reconstructed from repeated retrospective cases and incomplete board evidence. Founder Ken Kolence independently corroborates the financing context and his own management shortfall, but no contemporaneous board vote survives. Kolence, printed pp. 18–21
• Provenance and origin — [retrospective witness · moderate] Johnson names market understanding, a personality others will follow, and a record without side dealing as central founder tests. CHM, PDF pp. 19–20
• Mechanism and reasoning — [researcher inference · moderate] Integrity may reduce adverse-selection and governance risk; followership may expand recruiting capacity. Precommitting transition evidence is an analyst safeguard against loyalty, personality, or hindsight determining management change. CHM, PDF pp. 19–20 and 22–23
• Historical and vehicle context — [researcher inference · moderate] Early technical founders often created categories before professional management markets matured. Boards were small and investor involvement personal, making judgment powerful but informal. Boole, printed pp. 3–10
• When to use — [researcher inference · moderate] Analyst-created application checklist: Use in founder diligence and every financing or board review. Do not infer integrity from charisma, pedigree, or a warm introduction; do not replace a founder merely because scale creates discomfort. CHM oral history, PDF pp. 19–20 and 22–23
• How to apply it — [researcher inference · moderate] Analyst-created application checklist: (1) Reference specific prior choices; (2) test candor with adverse facts; (3) map who follows and why; (4) define the CEO's current job; (5) predefine scale milestones and failure signals; (6) name the board decision owner; (7) kill if material deception appears; (8) revisit role fit at each financing. CHM oral history, PDF pp. 19–20 and 22–23 · Bancroft, printed pp. 87–91
• Examples
  • Amgen — documented recruitment episode. [retrospective witness · moderate] Johnson recalls that he and Bowes checked Rathmann through 3M and Abbott contacts before recruiting him; he credits Bowes, Alstetter, Salser, Rathmann, Baddour, and others rather than a lone-investor story. Bancroft, printed pp. 42–46
  • Boole & Babbage — documented management transition and exception. [retrospective witness · moderate] Johnson says Kolence and Katch needed experienced management and accepts responsibility for at least one later poor executive choice; Kolence independently calls himself a poor president. The case supports transition need and shows that investor-led replacement can also fail. Bancroft, printed pp. 87–91 · Kolence, printed pp. 18–21
  • Tandem — researcher-applied leadership illustration. [researcher inference · moderate] Johnson praised Treybig’s market vision and execution; Byers independently identifies Treybig and Loustaunou, his own diligence, Perkins’s control, and Johnson’s board seat. The record supports leader quality and shared roles, not a Johnson-authored succession test. Dennis–Johnson, 02:23:57–02:24:29 · Byers, printed pp. 9–10/PDF pp. 18–19
• Evidence — [investor-stated · moderate] The criteria are direct. [retrospective witness · moderate] Bancroft and the Boole company interview support the management-change pattern; they do not preserve formal board votes or the counterfactual under unchanged leadership. Bancroft, printed pp. 87–91 · Boole, printed pp. 9–10
• Limits, failure modes, and exception ledger — [researcher inference · insufficient public record] “Followership” can reward similarity and charisma; “professionalize” can become founder displacement without causal proof. Attempted routes and results: 2019 CHM and 2008 Bancroft yielded Johnson’s criteria and later management recollections; the 1996 Kolence history corroborated founder/product context and Kolence’s self-assessed management failure; the Byers history constrained Tandem roles; the Amgen issuer archive did not expose private reference notes or board votes. Employment terms, option grants, and minutes remain missing. Amgen annual-report archive
• Evolution and contradictions — [investor-stated · moderate] Johnson's rule combines emotional commitment with willingness to replace management, yet his shutdown admission shows that loyalty sometimes overrode economic discipline. The tension is unresolved, not averaged away. Dennis–Johnson, 02:25:31–02:27:01
• Modern VC translation (as of 2026-08-21) — [researcher inference · moderate] [analyst-created hypothetical] Analyst-created current transfer hypothesis: document integrity references, recruiting evidence, role expectations, and transition triggers before investing. This is an analyst checklist, not a claim about present-day founder or board behavior. Hypothetical: a founder remains product chief after two missed enterprise quarters while a recruited commercial CEO owns sales; the board reassesses the split against prewritten milestones rather than prestige. CHM, PDF pp. 19–20 and 22–23 · Bancroft, printed pp. 87–91

4. Invest capital, time, judgment, and emotional capacity together

• Rule — [researcher inference · moderate] Treat a venture investment as a bundle of capital, advisory time, board judgment, recruiting help, and emotional commitment; invest only when the vehicle and partner have capacity for the promised work. Johnson states the contribution bundle; capacity underwriting is researcher synthesis. Dennis–Johnson, 02:25:31–02:27:01 and 02:29:07–02:29:58
• Evidence status — [researcher inference · moderate] The contribution bundle is investor-stated; the capacity condition is reconstructed. Named application evidence survives for Boole, Tandem, and Amgen, but hours, causal effects, and portfolio-wide consistency are not public. Bancroft, printed pp. 42–47 and 87–91
• Provenance and origin — [investor-stated · moderate] In the joint oral history Johnson says venture investors contribute money, consulting/time, and emotional involvement. Dennis–Johnson, 02:25:31–02:26:30
• Mechanism and reasoning — [researcher inference · moderate] Johnson’s cases combine product, hiring, financing, and governance work; the analyst hypothesis is that a trusted, accountable board participant can reduce coordination costs only when time and authority match the bottleneck. Bancroft, printed pp. 42–47 and 87–91 · Byers, printed pp. 9–10/PDF pp. 18–19
• Historical and vehicle context — [researcher inference · moderate] Johnson's personal/family capital avoided a large institutional portfolio, while early deals used small boards and weekly contact. This made high-touch work feasible but concentrated key-person risk and blurred paid labor, ownership, and personal loyalty. CHM, PDF pp. 22–23 and 27–29
• When to use — [researcher inference · moderate] Analyst-created application checklist: Use when board work can materially change recruiting, financing, product-market learning, or governance. Do not promise high-touch support when the investor lacks time, expertise, formal rights, or management trust. Dennis–Johnson, 02:25:31–02:27:01
• How to apply it — [researcher inference · moderate] Analyst-created application checklist: (1) Name the bottleneck; (2) assign founder, investor, board, and specialist ownership; (3) budget partner hours; (4) define rights and escalation; (5) set an observable outcome; (6) stop “help” that lacks a changed milestone; (7) kill the investment if success requires authority the investor will not have. Dennis–Johnson, 02:25:31–02:27:01 · Bancroft, printed pp. 87–91
• Examples
  • Boole & Babbage — documented board work and exception. [retrospective witness · moderate] Johnson describes weekly visits, executive recruiting, chairman work, and platform adaptation; he also assigns product insight to Kolence and Katch and admits poor management selection. Active involvement was real but not uniformly correct. CHM, PDF pp. 21–23 · Bancroft, printed pp. 87–91
  • Tandem — shared-role illustration. [retrospective witness · moderate] Byers sourced and diligenced, Perkins controlled the formation opportunity, Johnson took Asset Management's board seat, and Treybig led the company. The example supports role separation, not Johnson-only causal credit. Byers, printed pp. 9–10/PDF pp. 18–19
  • Amgen — researcher-applied contribution-bundle illustration. [researcher inference · moderate] Johnson reports family-partnership capital, board service, reference work, and recruiting with Bowes; he attributes execution to Rathmann and the company team. The record supports multiple contribution types but not their incremental causal value or Johnson-only credit. Bancroft, printed pp. 42–47
• Evidence — [investor-stated · moderate] The contribution bundle is direct testimony. [retrospective witness · moderate] Byers independently constrains the Tandem role map, while Boole records show both useful intervention and investor error. Dennis–Johnson, 02:25:31–02:27:01 · Byers, printed pp. 9–10/PDF pp. 18–19 · Boole, printed pp. 9–10
• Limits, failure modes, and exception ledger — [researcher inference · insufficient public record] Contact frequency is not causal value; emotion can prolong losses; boards can overreach; one partner cannot deeply serve an expanding portfolio. Attempted routes and results: CHM/Bancroft documented Johnson’s selected board work; Byers independently constrained Tandem roles; the 1997 Boole history supplied company-specific intervention; Amgen annual reports supplied issuer chronology but not investor time or causality. Time logs, minutes, recruiting records, and counterfactual outcomes remain unavailable. Amgen annual-report archive
• Evolution and contradictions — [investor-stated · moderate] Johnson contrasts company-building venture with later fee- and valuation-driven capital, yet he also says newer large financings can reduce a startup's immediate bankruptcy risk. His critique is about the work/capital mix, not proof that every large fund fails to help. CHM, PDF pp. 27–29
• Modern VC translation (as of 2026-08-21) — [researcher inference · moderate] [analyst-created hypothetical] Analyst-created current transfer hypothesis: price partner time and assign one accountable owner to a measurable bottleneck before approval. This is a capacity-planning hypothesis, not a claim that current boards or platform teams universally diffuse responsibility. Hypothetical: a fund budgets ten named-partner hours monthly for a regulated-fintech company's first bank integration and stops the intervention when contracting is no longer the bottleneck. Dennis–Johnson, 02:25:31–02:27:01 · Bancroft, printed pp. 87–91

5. Match capital intensity and duration to the actual vehicle

• Rule — [researcher inference · moderate] Reject or redesign opportunities whose required check, reserves, holding period, liquidity, governance, or geography exceeds the authority and duration of the investing vehicle. This combined rule is an analyst reconstruction from Johnson's separate capital-capacity and duration observations. Dennis–Johnson, 02:19:03–02:20:35 and 02:29:07–02:29:58 · Bancroft, printed pp. 88–91
• Evidence status — [researcher inference · moderate] Johnson states individual components—private/family capital, cleantech capacity limits, and long holding periods—but the integrated vehicle-fit rule is researcher-reconstructed. Dennis–Johnson, 02:19:03–02:20:35 and 02:29:07–02:29:58 · Bancroft, printed pp. 88–91
• Provenance and origin — [retrospective witness · moderate] Johnson distinguishes family/private capital from institution-backed venture and says his small family vehicle was unlikely to finance capital-heavy cleantech. Dennis–Johnson, 02:19:03–02:20:35 and 02:29:07–02:29:58
• Mechanism and reasoning — [researcher inference · moderate] A sound company can be a bad fit when the vehicle cannot fund milestones, preserve ownership, wait for liquidity, exercise governance, or survive key-person transition. The mismatch risks forced dilution, premature exit, or unproductive rescue capital; these consequences are analyst deductions, not Johnson-reported outcomes. Dennis–Johnson, 02:29:07–02:29:58 · Bancroft, printed pp. 88–91
• Historical and vehicle context — [contemporaneous record · strong] The Small Business Investment Act created the SBIC framework that D&J used, but the statute does not establish D&J's license, borrowing, compliance, or returns. [retrospective witness · moderate] Johnson separately remembers personal savings and a family loan as private formation capital; in a later bounded account he described family/individual-capital activity and said he and his wife were LPs behind a separately managed firm. The firm's legal identity, LP terms, manager economics, and decision rights remain unknown. Small Business Investment Act · CHM, PDF pp. 11–13 · CHM Blazing a Trail video, 00:53:48–00:54:12
• When to use — [researcher inference · moderate] Analyst-created application checklist: Use before every IC approval and reserve decision. Do not infer fit from the labels “evergreen,” “family office,” or “venture fund”; inspect the governing agreement, capital calls, decision rights, liquidity needs, and key-person provisions. Dennis–Johnson, 02:19:03–02:20:35 and 02:29:07–02:29:58
• How to apply it — [researcher inference · moderate] Analyst-created application checklist: (1) Model capital to the next two value inflections; (2) set target and minimum ownership; (3) stress dilution and reserves; (4) map duration and liquidity; (5) map decision/geographic authority; (6) test compensation and stakeholder incentives; (7) kill or syndicate when a downside case exceeds capacity. Dennis–Johnson, 02:19:03–02:20:35 and 02:29:07–02:29:58 · Bancroft, printed pp. 88–91
• Examples
  • Draper & Johnson — vehicle-formation illustration. [retrospective witness · provisional] Johnson remembers contributing $75,000 toward $150,000 of private capital and using the SBIC form; government leverage amounts and legal records conflict or remain absent. CHM, PDF pp. 11–13 · Draper oral history, printed pp. 23–27
  • Boole & Babbage — duration illustration. [retrospective witness · moderate] Johnson says a roughly thirty-year hold was possible because he invested as an individual, not through a conventional finite venture fund, and later contrasted that private-capital patience with outside-LP distribution pressure; BMC-stock value subsequently fell, so headline sale value was not realized proceeds. Bancroft, printed pp. 88–91 · CHM Pioneers video, 00:53:12–00:54:19
  • Cleantech — exception/non-application. [investor-stated · moderate] Johnson viewed the market as important but said his small private-family vehicle could not supply the required tens of millions. The model's correct output was vehicle-level non-participation, not a sector-quality judgment. Dennis–Johnson, 02:29:07–02:29:58
• Evidence — [retrospective witness · moderate] Direct passages connect capital form to feasible strategy, including the speaker-checked Boole-duration and later LP-structure accounts. [contemporaneous record · strong] The statute supplies only regulatory context. No partnership agreement or audited vehicle record validates legal identity, LP terms, manager economics, decision rights, fees, carry, reserves, or performance. Small Business Investment Act · Bancroft, printed pp. 88–91 · CHM Pioneers video, 00:53:12–00:54:19 · CHM Blazing a Trail video, 00:53:48–00:54:12
• Limits, failure modes, and exception ledger — [researcher inference · insufficient public record] Patient capital can become undisciplined capital, and syndication can dilute accountability. Attempted routes: CHM/Bancroft/Draper histories, official AMC pages, the statute, Stanford Engineering's SEVF retrospective, and successor filings. Missing: D&J/AMC governing agreements, fees/carry, calls, ownership schedules, reserves, distributions, LP rights, succession, and complete regional-vehicle economics. Stanford Engineering Venture Fund retrospective
• Evolution and contradictions — [researcher inference · moderate] Johnson moved across SBIC, personal/family, named Asset Management entities, SEVF, and regional vehicles, but public naming is too inconsistent to infer one stable vehicle doctrine. Patience at Boole coexists with his admission that some weak companies should have been closed. CHM, PDF pp. 11–16 and 23 · Bancroft, printed pp. 88–91
• Modern VC translation (as of 2026-08-21) — [researcher inference · moderate] [analyst-created hypothetical] Analyst-created current transfer hypothesis: test whether a vehicle's documented capital, duration, rights, reserves, and key-person continuity cover the downside case. This is a checklist hypothesis, not a claim that every current fund uses or needs the same instruments. Hypothetical: a small seed fund passes on a capital-intensive fusion plant unless a strategic syndicate contractually maps later facility capital without breaching the seed fund's stated ownership floor. Dennis–Johnson, 02:19:03–02:20:35 and 02:29:07–02:29:58 · CHM Blazing a Trail video, 00:53:48–00:54:12

6. Preserve upside options, but precommit milestones and stop rules

• Rule — [researcher inference · moderate] Give a technically credible company enough time to reach explicit value inflections, while precommitting kill criteria so patience does not become repeated rescue financing. This option-and-stop formulation is an analyst synthesis of Johnson's patience and shutdown admissions. CHM, PDF p. 23 · Bancroft, printed pp. 88–91
• Evidence status — [researcher inference · provisional] Johnson supports patience and explicitly confesses weak shutdown discipline. Application depth remains below target: Boole and Talkway are identified, but the early-AI investment is not; Talkway's funding sequence lacks checks and milestones. The three bullets below are two named retrospective cases plus one honest route-specific identity shortfall, not three independently verified stop-rule applications. CHM, PDF pp. 20–23 · Bancroft, printed pp. 88–91 · CHM Innovation Is Nothing New video, 01:04:29–01:06:31
• Provenance and origin — [investor-stated · moderate] Johnson says some companies can take twenty years, but agrees with Craig Taylor's criticism that he was poor at stopping further money and effort. CHM, PDF p. 23
• Mechanism and reasoning — [researcher inference · provisional] Uncertain technical projects can create option value only if each tranche buys informative evidence. The milestone/abandonment mechanism is an analyst reconstruction; Johnson's record supplies long-duration experience and a contrary shutdown admission, not a documented tranche system. CHM, PDF p. 23 · Bancroft, printed pp. 88–91
• Historical and vehicle context — [researcher inference · moderate] Personal/family capital could wait beyond a finite fund life, as Boole illustrates. That flexibility is necessary for some long-cycle work but removes an external liquidation discipline. Bancroft, printed pp. 88–91
• When to use — [researcher inference · moderate] Analyst-created application checklist: Use for long-cycle science, new infrastructure, or market-creation bets where learning is measurable. Do not use when milestones are unfalsifiable, the next round only extends runway, or management repeatedly misses without a changed plan. CHM oral history, PDF pp. 20–23
• How to apply it — [researcher inference · moderate] Analyst-created application checklist: (1) Define the next irreversible uncertainty; (2) price the tranche needed to resolve it; (3) set date, evidence threshold, and owner; (4) define follow-on, pause, sale, or shutdown paths ex ante; (5) reassess independent of sunk cost; (6) report both capital and partner-time exposure. CHM oral history, PDF pp. 20–23 · Talkway failure, 01:04:29–01:06:31
• Examples
  • Boole & Babbage — conditional patience. [retrospective witness · moderate] The company survived platform shifts and management changes and eventually sold for BMC stock, but Johnson says the received stock later fell and accepts that guidance could have been better. Patience preserved an option; it does not prove an optimal return. Bancroft, printed pp. 88–91 · Boole, printed pp. 9–10
  • Early AI — timing failure. [retrospective witness · provisional] Johnson says the unidentified company was judged too early and failed to produce a result. Because check, follow-ons, and milestones are unknown, it cannot be evidence for staged financing or disciplined option sizing. CHM, PDF pp. 20–21
  • Talkway — funded-pivot failure/exception. [retrospective witness · moderate] Johnson says the investing group helped start the company and financed a pivot from failed advertising services to video email, but demand did not appear and he did not fully understand why it failed. The case shows continued funding without a successful demand signal; missing checks, dates, decision rights, and stop criteria prevent a staged-financing conclusion. Speaker-checked machine captions support paraphrase only. CHM Innovation Is Nothing New video, 01:04:29–01:06:31
• Evidence — [researcher inference · provisional] Johnson supplies both the patience thesis and its strongest self-critique. Boole and Talkway are named applications, but no independent source reconstructs Johnson's stop-rule process; the unidentified early-AI loss remains a boundary. Talkway adds a funded-pivot failure, not verified tranche discipline or complete causality. CHM, PDF pp. 20–23 · Bancroft, printed pp. 88–91 · CHM Innovation Is Nothing New video, 01:04:29–01:06:31
• Limits, failure modes, and exception ledger — [researcher inference · insufficient public record] Attempted routes and results: the 2019 CHM transcript yielded the shutdown admission, unidentified early-AI loss, and unnamed living-dead cohort; speaker-checking the full CHM video yielded Johnson's uninterrupted Talkway funding-and-pivot account; Bancroft and the 1997 Boole history yielded one named long hold; the living-dead article supplied a general concept, not a Johnson application; company, filing, and archive searches did not recover Talkway financing schedules or identify the early-AI company. Missing: a Johnson-authored reserve framework, checks, board minutes, loss ratios, stop decisions, and any documented later process correction. Living-dead scholarship · CHM Innovation Is Nothing New video, 01:04:29–01:06:31
• Evolution and contradictions — [investor-stated · moderate] Johnson continued to defend long holds while acknowledging that some weak companies should have been closed. The evidence supports a durable tension, not a documented correction to his process. CHM, PDF p. 23
• Modern VC translation (as of 2026-08-21) — [researcher inference · moderate] [analyst-created hypothetical] Analyst-created current transfer hypothesis: prompted by Talkway's reported funded pivot without demand, judge a continuation only by the incremental information it is designed to buy, with an ex-ante stop path. This is a decision-test hypothesis, not evidence that any current financing instrument systematically delays failure recognition. Hypothetical: a robotics company receives one six-month tranche to prove 99.9% task reliability at a named customer; missing the threshold without a validated alternative buyer triggers sale or shutdown review. CHM Innovation Is Nothing New video, 01:04:29–01:06:31

7. Use networks for access, then separate referral, diligence, capital, governance, and credit

• Rule — [researcher inference · moderate] Cultivate trusted referral and co-investor networks, but record who introduced, assessed, approved, financed, governed, recruited, and narrated each deal before assigning credit. The role-ledger requirement is an analyst reconstruction prompted by Johnson's network account and independently constrained Tandem testimony. CHM, PDF pp. 13–16 and 21–23 · Byers, printed pp. 9–10/PDF pp. 18–19
• Evidence status — [researcher inference · moderate] Johnson describes network sourcing directly; the role-separation rule is reconstructed to correct attribution errors visible across Boole, Amgen, Tandem, and Coherent. CHM, PDF pp. 13–16 and 21–23 · Bancroft, printed pp. 42–46 and 87–91 · Byers, printed pp. 9–10/PDF pp. 18–19
• Provenance and origin — [retrospective witness · moderate] Johnson says early sourcing included cold visits, newspapers, lawyers, accountants, friends, and other investors; sparse deal flow made networks useful but not exclusive. CHM, PDF pp. 13 and 15–16
• Mechanism and reasoning — [researcher inference · moderate] Trusted networks may reduce search and reference costs; the role ledger is an analyst control against converting warm access into false picking credit. Boole, Amgen, and Tandem show that referral, financing, diligence, governance, and recruiting can belong to different people. Bancroft, printed pp. 42–46 and 87–91 · Byers, printed pp. 9–10/PDF pp. 18–19
• Historical and vehicle context — [retrospective witness · moderate] Johnson remembers an early cooperative market where investors often admitted one another by phone; this access regime was small, local, and relationship-heavy. Dennis–Johnson, 02:20:35–02:21:32
• When to use — [researcher inference · moderate] Analyst-created application checklist: Use in sourcing analytics, reference work, syndication, and deal attribution. Do not treat referral count as decision quality or a board seat as proof of sourcing, ownership, or causal influence. Bancroft, printed pp. 42–46 and 87–91 · Byers, printed pp. 9–10
• How to apply it — [researcher inference · moderate] Analyst-created application checklist: (1) Log the originating person and route; (2) distinguish sponsor from referrer; (3) assign diligence modules; (4) record legal approval and capital vehicle; (5) record negotiated rights; (6) map board/recruiting/follow-on/exit work; (7) retain unknown where evidence is absent; (8) test whether access is repeatable beyond one network. Bancroft, printed pp. 42–46 and 87–91 · Byers, printed pp. 9–10
• Examples
  • Boole & Babbage — referral/finance/governance split. [retrospective witness · moderate] John Bryan referred Kolence and Katch; Johnson, Bryan, Bill Edwards, and others participated in or organized financing; Johnson later chaired and helped recruit management. The record does not support simplifying all roles to “Johnson led the deal.” CHM, PDF pp. 21–23 · Bancroft, printed pp. 62 and 87–91
  • Amgen — invitation/formation/recruiting split. [retrospective witness · moderate] Bowes and Alstetter formed the corporation; Salser supplied the forming scientific role; Bowes invited Johnson; Bowes and Johnson recruited Rathmann; Johnson reports a small family-partnership investment. Ownership and cash-flow evidence is missing. Bancroft, printed pp. 42–46
  • Coherent — exception/uncertainty boundary. [retrospective witness · insufficient public record] Johnson calls it an early independent deal but explicitly says he does not remember how he heard about it; his later Rockefeller relationship appears connected to a Coherent director, not proof that Rockefeller sourced Coherent. CHM, PDF pp. 15–16
• Evidence — [retrospective witness · moderate] The three examples expose distinct roles and memory gaps. [retrospective witness · moderate] Byers independently shows the Tandem split—Byers diligence, Perkins control, Johnson board seat—which supports the role-ledger method beyond Johnson's own narration. Byers, printed pp. 9–10/PDF pp. 18–19
• Limits, failure modes, and exception ledger — [researcher inference · insufficient public record] Networks can entrench homophily, hide rejected opportunities, and launder later credit. Attempted routes and results: CHM and Bancroft founder/investor histories yielded Boole, Amgen, Coherent, and Intel role fragments; Byers independently constrained Tandem diligence, syndicate control, and board assignment; Kolence independently described Boole's product and his management weakness; official Amgen biographies and annual reports yielded issuer chronology; the IDEC SEC filing yielded issuer ownership history but no Johnson allocation. No referral log, IC vote, allocation schedule, side letter, board minutes, or complete pass ledger surfaced, so sourcing hit rate and individual alpha remain uncomputable. Amgen annual-report archive · IDEC 2004 Form 10-K
• Evolution and contradictions — [retrospective witness · moderate] Johnson says early venture was open and cooperative, while the Intel story shows access could still be controlled by a syndicate lead. Network openness was conditional, not universal. CHM, PDF pp. 14–15
• Modern VC translation (as of 2026-08-21) — [researcher inference · moderate] [analyst-created hypothetical] Analyst-created current transfer hypothesis: preserve warm-reference efficiency while measuring route diversity and role-specific contribution. This is an attribution hypothesis, not a claim about current carry or credit practices. Hypothetical: an IC memo credits a university scientist with referral, a principal with technical diligence, a partner with negotiation, and an operator with CEO recruiting; any carry attribution follows a preexisting written policy. Bancroft, printed pp. 42–46 and 87–91 · Byers, printed pp. 9–10

8. Audit power-law stories against ownership, cash flows, and the full denominator

• Rule — [researcher inference · moderate] Evaluate venture results at both company and security levels; show ownership, dilution, capital, timing, realized value, gross/net basis, and results with and without the dominant outlier before claiming skill. The audit rule is researcher-created from Johnson's informal distribution and Amgen/Boole realization warnings. CHM, PDF pp. 18–19 · Bancroft, printed pp. 90–91
• Evidence status — [researcher inference · provisional] Johnson directly describes an informal skewed distribution and Amgen's outlier importance, but the full audit rule and security-level requirements are researcher-created. Amgen, Boole, and Teradyne provide three distinct retrospective illustrations; none supplies a complete, independently verified personal cash-flow chain. CHM, PDF pp. 18–19 · Bancroft, printed pp. 90–91
• Provenance and origin — [investor-stated · provisional] Johnson offers an explicitly non-mathematical ten-deal pattern—roughly one exceptional outcome, several solid outcomes, several failures, and one or two living dead. CHM, PDF p. 19
• Mechanism and reasoning — [researcher inference · strong] Company fame does not determine investor return. Security-level outcomes depend on check, ownership, dilution, follow-ons, timing, terms, and realized proceeds; Johnson's Boole stock-account and Amgen outlier comment illustrate why, while missing companies and passes create survivorship bias. Bancroft, printed pp. 90–91 · CHM, PDF pp. 18–19
• Historical and vehicle context — [researcher inference · moderate] Johnson's record spans personal/family capital, D&J's SBIC, multiple Asset Management entities, SEVF, regional vehicles, and board or invitation roles. Combining them into one portfolio would mix owners, incentives, periods, and return bases. source-map vehicle boundary
• When to use — [researcher inference · strong] Analyst-created application checklist: Use for every track-record, Pareto, hit-rate, and peer comparison. Do not infer return from exit value, public-company success, board service, or a promotional “notable investments” list. CHM oral history, PDF pp. 18–19 · Bancroft, printed pp. 90–91
• How to apply it — [researcher inference · strong] Analyst-created application checklist: (1) Define entity/vehicle/time scope; (2) deduplicate follow-ons and aliases; (3) reconcile known, searched, researched, reconstructed, unresolved, and excluded states; (4) collect cash flows and ownership; (5) compute only supported gross/net metrics; (6) show ex-top-one and ordinary outcomes; (7) label all reported figures and missing terms. 37-family reconciliation · CHM oral history, PDF pp. 18–19
• Examples
  • Amgen — dominant-outlier exception/boundary. [investor-stated · moderate] Johnson says Amgen carried much of his record and jokes he would have been better off investing then fishing for twenty years. That supports outlier dependence, not a personal return multiple; the first investment is only retrospectively remembered and family-partnership economics are unverified. CHM, PDF pp. 18–19 · Bancroft, printed pp. 42–46
  • Boole & Babbage — realization warning. [retrospective witness · moderate] Johnson says the BMC consideration was stock and later fell; he rejects the comforting suggestion that investors realized a billion dollars. Company transaction value was not realized investor cash. Bancroft, printed pp. 90–91
  • Teradyne — distinct realization illustration. [retrospective witness · provisional] Johnson recalls buying founder Alex d'Arbeloff's personal shares, serving on the board for roughly a decade, and doing well after the company went public. The account establishes a different company and possible liquidity route, but not the price, ownership, dilution, sale timing, cash flows, or an independent security chain. CHM, PDF p. 18
• Evidence — [researcher inference · provisional] Johnson's outcome distribution is explicitly impressionistic. Amgen, Boole, and Teradyne are distinct applications of outlier, realization, and missing-security controls, but all rely materially on Johnson's retrospective narration; independent application depth and every complete person-to-vehicle cash-flow chain remain below target. CHM, PDF pp. 18–19 · Bancroft, printed pp. 90–91
• Limits, failure modes, and exception ledger — [researcher inference · insufficient public record] A 37-family discovery set is not a funded portfolio and is dwarfed by promotional totals above 200 or 250. Attempted routes and results: AMC Notable Investments yielded a survivor-selected list; SEC IDEC filings and the Amgen annual-report archive yielded issuer records but no Johnson security chain; Stanford's SEVF retrospective yielded vehicle-level claims but not holdings; regional reconstruction reports yielded program claims but not reconciled economics; CourtListener's Pitch Johnson query yielded no usable investment-schedule or cash-flow record. Missing: allocation, ownership, dilution, fees/carry, calls, distributions, write-offs, passes, and audited vehicle returns. AMC notable investments · IDEC filing · Amgen archive · Stanford SEVF · CourtListener query · 37-family deal ledger
• Evolution and contradictions — [investor-stated · moderate] Johnson celebrates company-building and public benefit while also acknowledging money as a motivation and criticizing fee/valuation-driven scale. Those are different objectives; without cash flows, the record cannot rank social contribution, company success, and investment performance on one axis. Dennis–Johnson, 02:24:29–02:24:58 · CHM, PDF pp. 27–29
• Modern VC translation (as of 2026-08-21) — [researcher inference · moderate] [analyst-created hypothetical] Analyst-created current transfer hypothesis: reconcile entity, ownership, cash flows, gross/net basis, and ex-outlier performance before crediting skill. This is an audit hypothesis, not a claim that particular current vehicles or marks are misleading. Hypothetical: present DPI, TVPI, ownership/dilution, and capital-weighted loss beside performance excluding the largest AI mark, with every affiliated SPV or transfer reconciled. 37-family reconciliation · CHM, PDF pp. 18–19

Evidence gaps and candidate models

• Pricing discipline as a standalone model — excluded. [researcher inference · insufficient public record] Johnson discusses markets and future value, but no consistent valuation ceiling, security preference, ownership target, or price-based pass rule survives across the 37-family roster. Attempted Tanous-book preview, oral histories, issuer filings, and company records did not establish a reusable pricing rule. 37-family deal ledger
• Small first checks followed by milestone-based follow-ons — excluded. [researcher inference · insufficient public record] Amgen includes a retrospectively remembered small family-partnership investment and later financing, but one case does not establish a general rule. The unidentified early-AI loss contains no public check or follow-on evidence and is not used to support this candidate. Bancroft, printed pp. 42–46 · CHM, PDF pp. 20–21
• VisiCorp and Talkway speaker-check boundary — included with bounded support. [retrospective witness · moderate] Speaker-checking Johnson's uninterrupted CHM video answer establishes VisiCorp's reported product-control/platform-transition miss and Talkway's reported funded pivot without demand. Machine captions support paraphrase only; neither segment supplies contemporaneous terms, board records, cash flows, or a complete causal proof. Both are used above as counterexamples, not standalone proof of a general rule. CHM Innovation Is Nothing New video, 00:55:20–01:06:31
• Policy advocacy as person-level investment edge — excluded. [researcher inference · insufficient public record] The 1978 statute and 1979 rule are contemporaneous outcomes; they do not isolate Johnson's causal contribution from the NVCA, lawmakers, agencies, peers, or the economy. Revenue Act of 1978 · 44 FR 37221
• Remaining retrieval priority — [researcher inference · insufficient public record]. Obtain D&J/AMC agreements, investment schedules, ownership and cash-flow records, the original Entrepreneurial Climate, lawful Tanous chapter access, Stanford course cases, company board records, and raw Something Ventured materials. These could validate, split, merge, or overturn the eight models. source-map unresolved routes · Something Ventured archive record

What the evidence supports

The durable conclusion is narrower than a hero narrative. Johnson developed a technically literate, people-centered, market-oriented practice and repeatedly contributed capital, recruiting, board work, and patient judgment to early technology companies. The record supports those behaviors through his oral histories and through participant accounts; it does not show that they generated repeatable, benchmark-beating returns. [researcher inference · moderate] Johnson, CHM oral history, PDF pp. 18–25 · Johnson, Bancroft oral history, printed pp. 39–63 · Byers, printed pp. 8–11

The most defensible edge thesis is therefore conditional: from the mid-1960s through early biotechnology, Johnson may have added value when scarce technical capital, local networks, complementary specialists, and long-duration governance could turn an incomplete technical opportunity into a company. The mechanism predicts traceable sourcing, a market test, explicit role division, recruiting or governance intervention, and patient capital in cases beyond the largest winner. It does not predict lone-investor authorship. [researcher inference · moderate] profile thesis and falsifiers · Tandem role evidence, Perkins printed pp. 38–43 · Boole oral history, printed pp. 4–10

Three findings survive skeptical review. First, Johnson's own record consistently separates technology from a served market and emphasizes people, integrity, and learning. [investor-stated · moderate] CHM, PDF pp. 8–10 and 19–20 Second, Tandem demonstrates useful role specialization: Perkins and Treybig formed the opportunity, Byers performed much of the diligence, and Johnson took Asset Management's board seat. [retrospective witness · moderate] Perkins, printed pp. 38–43 · Byers, printed pp. 9–11 Third, Johnson openly identified errors—insufficient capital for Intel, inadequate pursuit of Apple, premature timing in an unidentified AI company, and excessive patience with weak companies—which gives the philosophy adverse cases rather than only winners. [investor-stated · moderate] CHM, PDF pp. 14–16 and 20–24

Amgen remains the strongest supporting case and the strongest reason not to overclaim. Johnson recalls a small investment through a family partnership, participation in recruiting George Rathmann, and years of board work; he also says Amgen could have carried him economically for decades. Bancroft, printed pp. 47–63 · CHM, PDF pp. 19–21 [retrospective witness · moderate]

The 2005 video's machine captions render a reported 2,685-to-1 figure at 00:55:31–00:55:46, but the audio was not independently verified and the metric, numerator, denominator, ownership, dilution, distributions, taxes, and exact wording are undefined. The figure is not used as a verified personal return. 2005 video, 00:55:31–00:55:46 [investor-stated · provisional]

Bill Bowes, Winston Salser, Sam Alstetter, Sam Wohlstadter, Rathmann, scientists, employees, directors, and other capital providers were essential; Johnson and Caruthers remember formation and recruiting roles differently. Bancroft, printed pp. 47–63 · Caruthers participant history, “During the time…” and “In February 1980…” paragraphs [retrospective witness · moderate]

What could falsify the thesis

The favorable hypothesis would become materially stronger if contemporaneous files showed the same sequence across ordinary cases: Johnson or his organization originated or independently validated an opportunity; wrote a market-and-people thesis before investment; secured identifiable ownership on known terms; caused a documented recruiting, governance, or financing change; and produced attractive realized cash flows after fees and losses. It would be stronger still if that pattern survived removal of Amgen. [researcher inference · strong] specified evidence test

It would be falsified or sharply downgraded by any of the following:

• Investment schedules and capital accounts showing that Amgen supplied nearly all realized value while the rest of the funded portfolio lost money or failed to clear a relevant benchmark. [researcher inference · strong] Amgen concentration admission, CHM PDF pp. 19–21
• Contemporaneous memos showing that partners, founders, or syndicate leads supplied the decisive sourcing, picking, recruiting, and governance work attributed retrospectively to Johnson. [researcher inference · strong] people and decision-rights audit
• A complete pass and loss record showing no repeatable relationship between the proposed market/people/governance process and outcomes outside the winning anecdotes. [researcher inference · strong] searched-denominator boundary
• Vehicle agreements showing that time horizon, reserves, approval rights, or incentives prevented the patient intervention implied by the philosophy. [researcher inference · moderate] vehicle audit
• Evidence that the apparent edge disappears after controlling for early-Valley access, SBIC leverage, public research, syndicate selection, and unusually favorable category timing. [researcher inference · moderate] national-system history · Small Business Investment Act of 1958, §§301–303

The current public record neither proves nor falsifies the return claim because it lacks the cash-flow and decision denominator. That is a result, not an invitation to substitute company fame, board tenure, or remembered lessons for investment economics. [researcher inference · strong] portfolio reconciliation

Attribution and denominator audit

The reconciled research roster contains 37 distinct company, pass, failure, role, and program families after the speaker-checked VisiCorp and Talkway cases are included. All 37 were searched and claim-researched at their documented public-access boundaries; zero have a complete target-person chain from legal investor and security through ownership, dilution, follow-ons, distributions, and realized proceeds. The 37-family research roster is not a funded-investment denominator, and a career coverage fraction or literal 80/20 result is not computable. [researcher inference · strong] source-map roster and boundary · deal reconciliation

The unidentified early-AI company stays in the 37-family roster because Johnson explicitly says he invested, even though its name, date, vehicle, product, terms, and loss mechanics remain unknown. It can support only a timing-error observation, not a rule about small checks, staging, reserves, or follow-ons. [investor-stated · provisional] CHM, PDF pp. 20–21

Attribution is case-specific. Boole begins with Ken Kolence and Dave Katch's product insight and John Bryan's referral; Johnson participated in organizing and financing and later governance, while Kolence's recollection of lead status and financing terms remains uncorroborated. [retrospective witness · provisional] Kolence oral history, printed pp. 18–23 · Bancroft, printed pp. 62–63 and 87–91 Tandem divides formation, diligence, capital, and the board seat among Perkins, Treybig, Byers, and Johnson. [retrospective witness · moderate] Perkins, printed pp. 38–43 · Byers, printed pp. 9–11 Amgen divides formation, science, sponsorship, recruitment, capital, operations, and governance across Bowes, Salser, Alstetter, Rathmann, Johnson, and others; the Caruthers and Johnson recruitment memories are preserved rather than forced into one story. [retrospective witness · moderate] Caruthers participant history · Bancroft, printed pp. 47–63

Coherent is retained with an explicit sourcing-direction unknown because Johnson says his Rockefeller chronology may be mixed up. VisiCorp and Talkway are target-specific funded-failure families: Dennis directly addresses Johnson, Johnson's turn continues without a handoff through both accounts, and Dennis explicitly resumes at 01:06:31. VisiCorp supports an ignored product-control warning and platform-transition miss; Talkway supports a funded pivot followed by absent demand and incomplete causal understanding. Machine captions support bounded paraphrase, not verbatim quotation, and neither case supplies a complete vehicle, board, ownership, shutdown, or cash-flow chain. [researcher inference · moderate] Coherent account, CHM PDF pp. 15–17 · Innovation video, 00:55:20–01:06:31 · speaker-control note

Transfer limits for modern VCs

The transferable practice is a diligence and governance discipline, not Johnson's historical portfolio construction. A modern analyst can test whether a technical team names a served market, whether specialists cover genuine knowledge gaps, whether referral and decision rights are separated, and whether governance milestones are written before patience becomes inertia. These are analyst-created applications supported by the historical behavior, not Johnson-authored contemporary rules. [researcher inference · moderate] modern translation · mental-model inventory

Several historical advantages do not transfer cleanly: scarce local capital, SBIC leverage, a small co-investor network, relatively direct board influence, family-capital flexibility, and early category formation. Today's larger funds, competitive rounds, global syndicates, secondary liquidity, platform organizations, faster technical cycles, and different regulatory risks change ownership, reserves, governance leverage, and time-to-proof. The package therefore frames every current application as an analyst hypothesis and requires current premises to be independently tested. [researcher inference · moderate] historical context and modern boundary · vehicle comparison

The most dangerous misuse is to convert “patient capital” into indefinite follow-ons or “learn the science” into false technical certainty. Johnson's own shutdown weakness and the unidentified early-AI loss are counterexamples. A usable modern checklist needs precommitted kill criteria, next-check milestones, ownership and dilution math, reserve limits, decision owners, and a record of passes and ordinary losses. [researcher inference · moderate] decision exercises · CHM, PDF pp. 20–24

Unresolved evidence

The largest gap is economic: no complete D&J or Asset Management investment schedule, partnership agreement, audited fund return, capital account, cap table, distribution history, write-off schedule, reserve policy, or ex-Amgen cash-flow series was located publicly. [researcher inference · insufficient public record] vehicle and retrieval ledger

The largest decision gap is contemporaneous evidence: investment memoranda, pass logs, board minutes, recruiting records, and follow-on votes would distinguish ex-ante judgment from retrospective coherence. The largest attribution gaps are legal investor identity and exact roles at Amgen, Applied Biosystems, BioSurface, Electroglas, Coherent, and the unidentified failure cases. [researcher inference · insufficient public record] source routes and barriers

Direct-voice gaps remain bounded rather than silently filled. The original Entrepreneurial Climate text, several other event recordings or full programs, the inaccessible 1983 Amgen-report route, and the broken putative Treybig transcript route were not available after the documented legitimate routes. By contrast, the 1998 Stanford/CISAC paper and the Pioneers, Innovation, and Blazing caption tracks were recovered and processed in the final audit. Missing bodies still cannot support unseen words or specific decisions. [researcher inference · strong] written-record boundary · spoken-record boundary · source-map archive routes

Central-claim citation audit trail

Ordinary-claim citation audit trail

The sample uses at least one claim from every pre-review core document and alternates biography, access, argument, talk, deal, philosophy, and model claims rather than selecting one convenient cluster.

Twenty-point acceptance evidence

The package passes as a bounded research product, not as proof of a track record. Its strongest contribution is an attribution-aware decision framework; its most important conclusion is that Johnson's repeatable investment alpha and portfolio economics remain unproved in the public record. central-claim audit · denominator reconciliation [researcher inference · strong]

Writings and direct written record

• ★★★ | Entrepreneurship and Democracy | https://www.hoover.org/research/entrepreneurship-and-democracy | speech 2004-10-08; adapted essay 2005 | full investor-authored Hoover essay; reliability A; heading/paragraph locators | Direct policy philosophy on markets, law, education, venture capital, economic freedom, and government; distinguish normative claims from proven outcomes.
• ★★★ | Promoting Innovation in the USA | https://obamawhitehouse.archives.gov/sites/default/files/microsites/ostp/compiled_rfi_responses_redacted.pdf | 2014-09-15 | full signed response in official White House RFI compilation; reliability A; printed pp. 662–665 | Late direct statement on science versus technology, markets, regulation, taxes, education, immigration, patience, and SBIC design.
• ★★★ | Health-Care Entrepreneurship in Poland | https://studylib.net/doc/14809333/ | delivered 2015-05-20 | full nine-page conference paper mirror; reliability B for hosted body pending event/original-file authentication | Direct, dated application of Johnson's market/technology/team/government framework to genomics, data, cell therapy, and Poland.
• ★★★ | Introduction to Tom Perkins oral history | https://digicoll.lib.berkeley.edu/record/103633/files/perkins_tom_donated.pdf | signed 2001-12-30; oral history published 2002 | full signed introduction in Bancroft/NVCA transcript; reliability A; printed pp. i–ii | Johnson's compact written assessment of incubation, governance, marketing, LP relations, and Perkins's influence; preserves admiration and attribution bias.
• ★★ | The Entrepreneurial Climate | https://obamawhitehouse.archives.gov/sites/default/files/microsites/ostp/compiled_rfi_responses_redacted.pdf | Johnson dates paper 1984; other profiles say 1982 | original body unavailable; title/seven elements restated by author on printed p. 662; reliability A for self-description, partial access | Exact-title/author/date searches across HBS/Baker, Stanford, Google Books, and web archives did not expose the original; content analysis is limited to Johnson's 2014 recap and is not conflated with the resolved 1998 Russia paper.
• ★★ | The Masters of Private Equity and Venture Capital — Pitch Johnson chapter | https://books.google.com/books/about/The_Masters_of_Private_Equity_and_Ventur.html?id=SYeduhx-hKAC | 2010 | Peter J. Tanous book; participant interview/chapter; reliability B; limited preview | The lawful preview was processed to its visible boundary; only displayed Johnson-attributed material may support claims, while the licensed remainder stays excluded from analysis.
• ★★ | The Application of Venture Capital and the Entrepreneurial Revolution in Russia — 1998 Stanford paper identity resolved | https://www.files.ethz.ch/isn/19993/Application_Venture_Capital.pdf | paper delivered 1998-06-05; edition dated 1998-08 | full Johnson-authored CISAC paper; reliability A; title/foreword printed pp. 3–7, argument pp. 7–15, author note p. 17 | Resolves the formerly untitled Stanford-attributed 1998 route; this is the single counted entry for the underlying work, which later context sections cite without rostering again.
• ★★ | HBS Alumni Achievement Award profile | https://www.alumni.hbs.edu/Documents/Awards/2012_HBS_AAA_Johnson.pdf | 2012 | institutional award profile with direct fragments; reliability B; p. 7 | Concise direct account of D&J sourcing and post-sale chronology; promotional context and 1967 formation date need controls.
• ★★ | A founding father of Silicon Valley venture capital guides entrepreneurs | https://www.alumni.hbs.edu/stories/Pages/story-bulletin.aspx?num=3351 | 2014-04-24 | HBS interview/profile with direct fragments; reliability B; paragraph locators | Late summary of sector focus, sourcing, portfolio-count claim, and teaching; useful for evolution and inconsistency checks.
• ★★ | Venture Capital Pioneer Illuminates the Silicon Valley Ecosystem | https://www.gsb.stanford.edu/insights/venture-capital-pioneer-illuminates-silicon-valley-ecosystem | 2013-01-16 | Stanford GSB written Q&A/interview; reliability B; question-heading locators | Direct answers on founder/market selection, impatience, agreements, future technologies, teaching, and Valley conditions.

Oral histories, talks, video, and interviews

• ★★★ | Franklin Pitcher “Pitch” Johnson — Bancroft oral history catalog | https://digicoll.lib.berkeley.edu/record/103604 | interviews 2008; published 2009 | official catalog and rights record; reliability A for custody/metadata; transcript plus six video discs | Canonical collection route and quotation-rights boundary; collapse with the transcript as one interview family.
• ★★★ | Franklin Pitcher “Pitch” Johnson — Bancroft oral history transcript | https://digicoll.lib.berkeley.edu/record/103604/files/johnson_franklin.pdf | three interviews in 2008; published 2009 | full 101-page participant transcript; reliability A for attributed speech/B for retrospective memory; quotation restricted | Core direct record for life, vehicles, Amgen, philosophy, policy, teaching, Eastern Europe, failures, and later questions.
• ★★★ | Oral History of Pitch Johnson | https://archive.computerhistory.org/resources/access/text/2020/05/102795131-05-01-acc.pdf | recorded 2019-09-12; released 2020 | full CHM transcript; reliability A for attributed speech/B for memory; 46 PDF pages | Late cross-check on formation capital, deals, misses, company-building, teaching, civic work, and critique of financialized venture.
• ★★★ | Oral History of Franklin Pitcher Johnson — Boole & Babbage | https://archive.computerhistory.org/resources/access/text/2015/06/102702043-05-01-acc.pdf | recorded 1997; CHM release later | full 12-page CHM participant transcript; reliability A/B | Deal-specific direct account of the software company's financing, board work, management transitions, market, and exit.
• ★★★ | Reid Dennis and Pitch Johnson — Venture Voices | https://digicoll.lib.berkeley.edu/record/218377/files/dennis_johnson_donated.pdf | interviewed 2008-09-22; published 2009 | full NVCA/Bancroft joint transcript; reliability A for attributed speech/B for memory; timestamp and printed-page locators; quotation restricted | Compact independent interview family on field history, people/markets, financial/time/emotional contribution, founder replacement, and failures.
• ★★★ | Innovation Is Nothing New — full speaker-controlled video | https://www.youtube.com/watch?v=gC6qJ7HuCX8 | recorded 2005-04-27; uploaded 2019 | full 1:39:48 official CHM video; machine-generated en-orig captions processed; reliability A for custody/bounded speaker attribution and B for caption wording/retrospective memory | Dennis asks Johnson about a bad deal at 00:55:20 and Dennis resumes at 01:06:31. At 00:55:31–00:55:46 the auto-captions render a Johnson-reported 2,685-to-1 Amgen figure; audio was not independently verified, and metric, numerator, denominator, ownership, dilution, distributions, and taxes are undefined, so it is not a verified return or calculation input. [investor-stated · provisional] VisiCorp is supported at 00:58:58–01:01:36 and Talkway at 01:04:29–01:06:31; captions support timestamped paraphrase, not verbatim quotation.
• ★★ | Innovation Is Nothing New — event record | https://computerhistory.org/events/innovation-nothing-new-100odd-years/ | 2005-04-27 | official CHM metadata; reliability A for date/participants | Canonical event control; count with the YouTube video as one appearance family.
• ★★★ | Pioneers of Venture Capital | https://www.youtube.com/watch?v=ult7VFgxiKo | 2002-09-30 | full official CHM video; machine-generated en-orig captions processed; reliability A for custody/bounded speaker attribution and B for caption wording/retrospective memory | Johnson turns are controlled by direct handoff at 00:29:50–00:33:20, 00:53:12–00:54:19, and 01:18:15–01:20:52. At 00:53:12–00:54:19 he reports Boole & Babbage as his first very valuable deal, says it took roughly 30 years, and contrasts private-capital patience with outside-LP distribution pressure; check, legal vehicle, ownership, and cash flows remain unknown. Caption wording is paraphrased, not quoted. [investor-stated · provisional]
• ★★★ | Something Ventured — Pitch Johnson interview-file route | https://purl.stanford.edu/mb678nw9491 | interview circa 2009; film released 2011 | Stanford collection metadata states that digital interview transcripts exist, but no public item-level file is exposed; reliability A for custody/metadata; access restricted to archival research | The public catalog and item route were exhausted; the raw transcript remains available only through restricted Special Collections access and was not requested, so analysis stops at custody metadata and does not infer the interview from the edited film.
• ★★ | Something Ventured official study guide | https://zeitgeistfilms.com/sitelets/somethingventured/pdf.php | film 2011 | official distributor guide; reliability B | Participant, runtime, and legal-viewing context for the edited documentary; no substitute for raw speaker evidence.
• ★★ | Blazing a Trail: Venture Capital Pioneers Reflect on the Industry | https://www.youtube.com/watch?v=oFVhGsovzAM | event 2019-03-28 | full official CHM video; machine-generated en-orig captions processed; reliability A for custody/bounded speaker attribution and B for caption wording/retrospective memory | Speaker-controlled Johnson turns at 00:07:08–00:10:00, 00:53:48–00:54:12, and 00:56:38–00:57:56 cover Amgen credit, family capital, and early-AI market timing. At 00:53:48–00:54:12 he reports that activity remained based on family/individual capital and that he and his wife were LPs behind a separately managed firm; the legal entity, LP terms, manager economics, and decision rights are not disclosed. Caption wording is paraphrased, not quoted; later Mallaby program is excluded. [investor-stated · provisional]
• ★★ | Founders of Silicon Valley Venture Capital panel — GTS highlights | https://techcrunch.com/2009/04/03/highlights-from-this-years-global-technology-symposium/ | 2009-04-03 | contemporaneous collective-panel report; reliability B; partial appearance | Johnson with Draper, Dennis, and Ann Winblad; route for event/video search, not speaker-specific claims from collective prose.
• ★★ | Pitching to become the new Silicon Valley | https://www.rt.com/business/russia-skolkovo-potential-technology/ | 2010-10-15 | direct broadcast interview transcript/excerpts; reliability B for attributed statements | Ex-ante Russia market-size, export, corruption, domestic-capital, and institutional views; venue framing needs explicit treatment.
• ★★ | Entrepreneur and society — Moscow lecture report | https://ria.ru/20101013/285074841.html | 2010-10-13 | contemporaneous Russian-language report of public lecture; reliability B; full accessible report but no recording/transcript | Report body, exact-title searches, venue/date searches, and video searches were processed on 2026-08-21; paraphrase only the reporter-attributed claims and do not promote them to verbatim speech.
• ★★ | BioForum 2015 event/report | https://www.biotechnologia.pl/biotechnologia/biosektor-zakonczyl-swoje-swieto%2C15304 | 2015-05-22 | contemporaneous Polish event report; reliability B | Authenticates the Wroclaw appearance and helps validate the hosted paper; no claim that every paper line was delivered verbatim.
• ★ | MVS25 Pitch Johnson clip | https://www.linkedin.com/posts/context-vc_at-mvs25-legendary-investor-pitch-johnson-activity-7349056552995053568-Pldd | event/upload 2025 | host-posted short clip and post text; reliability B; access partial | The post, clip metadata, organizer/name variants, and exact-phrase searches were processed on 2026-08-21; no full public program surfaced, so only the host-posted excerpt is usable and it is not ranked as a full appearance.
• ★ | SIEPR Economic Summit — personalized medicine panel | https://siepr.stanford.edu/events/summit/siepr-economic-summit-2010 | 2010-03-12 | official event program; reliability A for metadata; body not located | Johnson moderated Brook Byers and Elizabeth Holmes; important adverse/retrospective context, but metadata does not prove his views or later knowledge.

Firm, vehicle, company, deal, and team records

• ★★★ | Pitch Johnson official AMC biography | https://amcfamilyoffice.com/pitch-johnson/ | live page accessed 2026-08-20 | official firm biography; reliability B/promotional; paragraph locators | 1962/1965 chronology, 250-plus portfolio claim, selected companies, boards, associations, teaching, and Eastern Europe.
• ★★★ | AMC About Us | https://amcfamilyoffice.com/about-us/ | live page accessed 2026-08-20; describes 2012 transition | official firm history; reliability B/promotional | Vehicle-era boundary, family-office transition, and direct “long view” fragment; current team is not the historical partnership.
• ★★★ | AMC Notable Investments | https://amcfamilyoffice.com/notable-investments/ | live page accessed 2026-08-20 | official firm portfolio/philosophy page; reliability B/promotional | Denominator lead and firm-stated large-market, management, and company-helping philosophy; visible page lacks a complete company list.
• ★★★ | Brook Byers oral history | https://digicoll.lib.berkeley.edu/record/103443/files/byers_brook.pdf | interviews 2002–2005; published 2006 | full Bancroft participant transcript; reliability A/B | First employee's account of Johnson, personal capital, Qume, Tandem diligence, board attribution, and apprenticeship.
• ★★★ | Ken Kolence oral history | https://archive.computerhistory.org/resources/text/Oral_History/Kolence_Ken/102702064.05.01.acc.pdf | interviewed 1996 | full CHM founder transcript; reliability A/B | Independent Boole & Babbage founder account of Johnson's lead role, reported financing terms, and software-market context.
• ★★★ | Tom Perkins oral history | https://digicoll.lib.berkeley.edu/record/103633/files/perkins_tom_donated.pdf | interviews 2001–2002 | full Bancroft/NVCA transcript; reliability A/B | Co-investor account of Tandem, Hybritech, Genentech exclusion, early biotech, and Johnson's role; also contains Johnson's signed introduction.
• ★★★ | Peter Crisp CHM oral history | https://archive.computerhistory.org/resources/access/text/2019/04/102717367-05-01-acc.pdf | recorded 2018; released 2019 | full participant transcript; reliability A/B | Independent witness that Johnson invited Venrock into Amgen and that Venrock passed; useful adverse comparator and attribution control.
• ★★★ | Amgen first annual report — exhausted public route | https://www.annualreports.com/HostedData/AnnualReportArchive/a/NASDAQ_AMGN_1983.pdf | 1983 report sought; route checked 2026-08-21 | legacy PDF now returns 404; no report body retrieved from the live host or public snapshot route; reliability A only if the missing issuer body is recovered | Processed to the access boundary: no 1983 operating, capitalization, board, risk, ownership, or Johnson-economics claim is drawn from this unavailable body.
• ★★★ | Amgen annual-report archive | https://investors.amgen.com/financials/annual-reports | current official archive checked 2026-08-21 | issuer archive; reliability A for hosted filings; visible archive begins in 2000 | Processed to the public boundary: the live archive does not expose the 1983 first report, so early-case evidence instead uses participant histories and keeps contemporaneous capitalization and Johnson cash flows unresolved.
• ★★★ | Tandem Computers oral history — James Treybig route exhausted | https://archive.computerhistory.org/resources/access/text/2013/04/102658223-05-01-acc.pdf | putative CHM route checked 2026-08-21 | URL returns 404 and the accession does not yield a public Treybig transcript in CHM/catalog or exact-title searches; access: unavailable | Processed to the access boundary; no founder quotation or term claim is imputed to this missing body. Founder-side context is controlled with Perkins printed pp. 38–43, Byers printed pp. 8–11, and Johnson’s 2019 transcript PDF pp. 18–19.
• ★★★ | The Chemical Synthesis of DNA/RNA: Our Gift to Science — Marvin Caruthers participant history | https://pmc.ncbi.nlm.nih.gov/articles/PMC3543024/ | 2013 | full first-person scientific/formation history; reliability A for Caruthers’s participation/B for retrospective memory; locator: paragraphs beginning “During the time…,” “In February 1980…,” and “We located Applied Biosystems…” | Replaces a misidentified CHM URL that resolves to John Sanguinetti, not Applied Biosystems; establishes the instrument-company fork, Lee Hood/Caruthers contribution, Sam Eletr recruitment, reported $3 million initial capital, customer deposits, first product timing, and attribution boundaries.
• ★★ | IDEC/Asset Management successor-role filing | https://www.sec.gov/Archives/edgar/data/949699/000094969905000059/pdf10k.pdf | filed 2005 | SEC 10-K; reliability A; director biography at p. 29 | Documents Craig Taylor, AMC Partners 89, Asset Management Associates 1989, and Alloy succession; a route to vehicle/entity reconciliation.
• ★★ | Stanford Engineering Venture Fund at twenty-five | https://engineering.stanford.edu/news/fearless-stanford-engineering-venture-fund-twenty-five | 2012 | official Stanford retrospective; reliability B | Vehicle strategy, volunteer team, reported 30%/36% annualized results, Johnson's SEVF II role, and explicit timing/luck control.
• ★★ | Renaissance Partners history | https://rp.com.pl/en/history/ | live successor page processed 2026-08-20/21 | full accessible company history; reliability C/promotional | Exact entity/Johnson/EBRD/IFC/fund-report searches established the regional manager route but exposed no Johnson-specific governing agreement, holding, fee/carry, or cash-flow body; no performance claim is drawn.
• ★★ | Stanford Bio-X donor story | https://biox.stanford.edu/node/1001 | date not stated | official Stanford institutional page; reliability B | Routes Johnson-family biotech philanthropy and research-network role; philanthropy and deal sourcing must remain distinct.
• ★ | Reconstruction Capital II reports | https://www.reconstructioncapital2.com/reports/ | archive processed 2026-08-21 | public company filing archive; reliability A for filed company facts; 2006 annual report title page and directors list at printed p. 3 | The 2006 audited report identifies Franklin Pitcher Johnson Jr. as a director of the Cayman limited-liability company; the archive was searched across later reports and provides no evidence of his personal investment, ownership, approval right, or attributed return.

Press, biography, witnesses, and critical leads

• ★★★ | William H. Draper III oral history | https://digicoll.lib.berkeley.edu/record/103600/files/draper_william.pdf | interviews 2008; published 2009 | full Bancroft participant transcript processed 2026-08-21; reliability A for attributed speech/B for retrospective memory; printed pp. 23–25 and 86–87 | Cofounder cross-check: Draper recalls $75,000 from each partner, a $300,000 ten-year 5% government loan, roughly 10–15 D&J investments, the Sutter Hill combination, and shared Electroglas recruiting/cash-crisis work; these remain witness claims pending SBIC and cash-flow records.
• ★★ | The Startup Game — Bill Draper | https://books.google.com/books/about/The_Startup_Game.html?id=h08NfXIK1fYC | 2011 | cofounder memoir; reliability B; partial preview/licensed body | Independent but retrospective D&J narrative; personal/family branding and later career must not be transferred to Johnson.
• ★★ | The Power Law — Sebastian Mallaby | https://books.google.com/books/about/The_Power_Law.html?id=FqVPEAAAQBAJ | 2022 | researched narrative history; reliability B; limited preview | Modern cross-investor narrative and endnote trail; use licensed body and underlying records for Pitch-specific claims.
• ★★ | VC: An American History — Tom Nicholas | https://books.google.com/books/about/VC.html?id=uSmWDwAAQBAJ | 2019 | scholarly history; reliability B; limited preview | SBIC, regional, vehicle, and base-rate context; trace person/deal claims to primary records.
• ★★ | Founders Field Guide / 25iq — A Dozen Things from Pitch Johnson | https://25iq.com/2016/06/11/a-dozen-things-ive-learned-from-pitch-johnson-about-venture-capital-and-startups/ | 2016-06-11 | synthesis with linked quotations; reliability C as discovery map | Efficient lead to original interviews and book passages; never treat the compilation as quote origin or independent corroboration.
• ★★ | Automatic Pilots | https://www.forbes.com/asap/2000/0529/148.html | 2000-05-29 | reported feature/interview; reliability B; access may be partial | Contemporary generational, governance, and lifestyle comparison with Dennis; exact Johnson fragments need full-body verification.
• ★★ | Buried Treasures | https://www.forbes.com/asap/1998/1005/067.html | 1998-10-05 | reported wealth profile; reliability B/C for estimates; partial access | Wealth and company-history lead only; never convert magazine estimates into audited personal or fund returns.
• ★★ | Real VCs help build great companies | https://www.edn.com/real-vcs-help-build-great-companies/ | 2000 | Don Valentine authored industry essay; reliability A for Valentine's words/B for industry description | Adversarial contemporary view of hands-on company building versus dot-com capital deployment; peer philosophy, not Johnson testimony.
• ★★ | Brook Byers — current official biography | https://www.kleinerperkins.com/people/brook-byers/ | live page accessed 2026-08-20 | current firm biography; reliability B/promotional | Cross-checks Byers’s career and portfolio claims; it does not establish his Johnson apprenticeship, which rests on the Bancroft oral history.
• ★ | Property Journal/WGN — Entrepreneurship and democracy | https://wgn.pl/artykuly/-/3703 | 2012 | Polish-language profile/interview; reliability C | Eastern Europe narrative and reported hit-rate lead; trace every number and quotation to Johnson's primary papers/interviews.

Policy, scholarship, historical context, and adverse controls

• ★★★ | Small Business Investment Act of 1958 | https://www.govinfo.gov/content/pkg/COMPS-1835/pdf/COMPS-1835.pdf | enacted 1958; current compilation | official statutory compilation; reliability A | Governing SBIC framework; statute does not prove D&J's license, draw, repayment, restriction compliance, or returns.
• ★★★ | Investment Duties regulation — 44 FR 37221 | https://archives.federalregister.gov/issue_slice/1979/6/26/37221-37229.pdf | 1979-06-26 | official final regulation; reliability A | Primary prudent-investor-rule date and text; controls compressed recollections and blocks sole-advocate credit.
• ★★★ | Revenue Act of 1978 | https://www.govinfo.gov/content/pkg/STATUTE-92/pdf/STATUTE-92-Pg2763.pdf | 1978-11-06 | enacted statute; reliability A | Capital-gains policy outcome; law establishes change, not Johnson's individual causal contribution.
• ★★★ | NVCA Board of Directors Handbook | https://nvca.org/wp-content/uploads/2023/05/2023-2024-NVCA-Board-of-Directors-Handbook.pdf | 2023–2024 | association record; reliability A for role chronology | Former-chair/director table and governance context; titles are not proof of authorship for collective lobbying wins.
• ★★ | The Economics of the Private Equity Market | https://www.federalreserve.gov/pubs/staffstudies/168/ss168.pdf | 1995 | Federal Reserve Staff Study 168; reliability A | Institutional history, incentives, information problems, fundraising, returns, and SBIC/LP context; not an AMC audit.
• ★★ | Financiers of Innovation | https://creativeclass.com/articles/Financiers%20of%20Innovation.pdf | 1997 | scholarly manuscript by Florida and Kenney; reliability B | West Coast network, The Group, SBIC, and industry formation; resolve names/dates against direct witnesses.
• ★★ | How Venture Capital Became a Component of the U.S. National System of Innovation | https://kenney.faculty.ucdavis.edu/wp-content/uploads/sites/332/2018/03/how-venture-capital-became-a-component-of.pdf | scholarly article | peer-reviewed institutional history; reliability B | Policy, regional, vehicle, and capital-supply context; prevents lone-pioneer causation.
• ★★ | Solving the problem of abundance | https://eprints.lse.ac.uk/125890/1/Solving_the_problem_of_abundance_venture_capital_and_the_making_of_asset-driven_inequalities.pdf | 2024/2025 | academic critical account; reliability B | Strongest located systemic critique: public research, SBIC support, tax/pension policy, inherited surplus, and asset inequality as rival explanation.
• ★★ | Founders in the Living-Dead Trap | https://doi.org/10.1515/erj-2015-0029 | 2016 | peer-reviewed theoretical article; reliability B | Investigates the “living dead” concept attributed to Johnson; trace phrase origin and test patience against escalation/closure bias.
• ★★ | GAO SBIC program reports — processed search and nearest period report | https://www.gao.gov/assets/rced-95-146fs.pdf | GAO/RCED-95-146FS, 1995 | full official fact sheet; reliability A; pp. 9–18, especially pp. 11–12 and figs. 1.4–1.9 | Search processed through exact program/date variants: the report documents 1990–1994 licensee counts, capital/leverage, and cumulative loan losses, but is too late to supply a 1962 D&J underwriting base rate; that earlier numerical benchmark remains unavailable rather than inferred.
• ★ | Asset Management/Johnson litigation and enforcement sweep | https://www.courtlistener.com/?q=%22Pitch%20Johnson%22&type=r&order_by=score%20desc | search processed 2026-08-20/21 | public case-law discovery route; reliability C for negative-search boundary | Exact-name, firm-alias, company, SEC and enforcement variants were screened; no verified target-person sanction surfaced, while false positives were excluded. Absence remains bounded, not proof.

Archive, catalog, and unresolved-record routes

• ★★ | Franklin P. “Pitch” Johnson papers — Stanford/OAC finding aid | https://oac.cdlib.org/findaid/ark:/13030/c8x06dp3/ | collection circa 1975–2008; finding aid processed 2026-08-21 | full public finding aid; reliability A for custody/scope; M1939, 78 linear feet/158 boxes | Identifies bulk Eastern European corporate/legal/financial material and company files for Amgen, Tandem, Hypres, and Biogen IDEC; item-level bodies are not public, and stated student/employee restrictions are respected.
• ★★ | Online Archive of California — Franklin P. Johnson search | https://oac.cdlib.org/search?query=%22Franklin%20P.%20Johnson%22 | catalog processed 2026-08-20/21 | archival union-catalog route; reliability A for metadata | Exact-name/firm/company variants surfaced the M1939 finding aid above and no second target-specific public deal collection; deduplicated rather than counted as a new body.
• ★★ | CHM oral-history and catalog search | https://www.computerhistory.org/collections/search/?q=%22Pitch%20Johnson%22 | catalog processed 2026-08-20/21 | official museum discovery route; reliability A for custody metadata | Johnson/name/company/event variants were enumerated and transcript/video manifestations deduplicated; inaccessible or unlabeled media remain non-evidence, and accessible transcripts are mapped separately above.
• ★★ | Smithsonian Something Ventured collection NMAH.AC.1467 | https://sova.si.edu/record/NMAH.AC.1467 | collection dates 2008–2010 | official finding aid; reliability A for custody/metadata | Routes raw Johnson interview assets and production records; collapse with Stanford manifestations by interview date.
• ★★ | Harvard Business School Baker Library search | https://id.lib.harvard.edu/alma/990020907640203941/catalog | catalog processed 2026-08-20/21 | library catalog; reliability A for displayed metadata; no lawful public full text exposed | Exact-title, author, course, and case variants were processed; the catalog did not expose The Entrepreneurial Climate or Johnson course/case bodies, so content analysis stops at authenticated metadata and Johnson's later recap.
• ★ | SBA FOIA and historical records | https://www.sba.gov/about-sba/open-government/foia | live route accessed 2026-08-20 | official records-request route; reliability A for process | D&J and Asset Management SBIC applications, licenses, debentures, reports, examinations, payoff, transfer, and enforcement are high-value nonpublic leads; no request sent.

Coverage assessment

Mechanical coverage

This SCOUT map contains 70 checked retrieval/manifestation routes: 28 ★★★ / 36 ★★ / 6 ★, or 10 written/direct-text routes; 17 oral-history, talk, video, and interview routes; 16 firm, vehicle, company, deal, and team routes; 10 press/biography/witness routes; 11 policy, scholarship, historical, and adverse-control routes; and 6 archive routes. After collapsing the Bancroft catalog/transcript pair, the Innovation event/video pair, and the three Something Ventured manifestations, those 70 routes represent 66 independent underlying works or work families. The recovered 1998 CISAC paper is counted once in the written record. All 70 route checkboxes are closed because the research phases processed each route to its stated access boundary; a checked unavailable or partial route does not imply that a full body was retrieved or create an additional independent work.

Name and identity sweep

Searches used Franklin Pitcher Johnson Jr., Franklin P. Johnson Jr., Franklin Johnson, F. P. Johnson, Pitch Johnson, Johnson, Franklin P., Draper & Johnson/Draper and Johnson, Asset Management Company, Asset Management Capital Company, Asset Management Associates/Partners/AMC Partners, and named colleagues, vehicles, companies, events, and countries. The target is separated from his father, Franklin P. “Pitch” Johnson Sr. (Stanford/Drake coach and 1924 Olympian), other Franklin Johnsons, Johnson & Johnson, generic asset-management firms, and later Asset Management Ventures or Alloy activity lacking target-specific evidence.

Direct voice and writings

The direct spine comprises the full three-session Bancroft history, the independent 2019 CHM interview, the 1997 Boole & Babbage interview, the joint Dennis–Johnson transcript, three official CHM videos processed through complete caption tracks and explicit speaker-turn controls, Johnson's full 1998 Russia paper, the access-restricted Something Ventured raw-interview route, and bounded Russia/Poland appearances. The written denominator remains 10 priority/direct-text routes representing ten candidate works or work families: four full authored bodies (Hoover, White House RFI, Poland health care, and the identified 1998 Russia paper), one full signed introduction, the partial original The Entrepreneurial Climate route, one licensed/limited-preview chapter, and three institutionally edited Q&A/profile routes. Metadata-only records are not analyzed as prose.

Bancroft and NVCA reserve publication rights; their transcripts are available for research and should be paraphrased rather than quoted. Stanford raw documentary materials carry repository/fair-use constraints. No full transcript or copyrighted article will be committed.

Vehicle and denominator boundary

At least ten legal/economic buckets require separation: Johnson personal/family capital; D&J SBIC; retained/assigned D&J positions; post-sale personal positions; the family SBIC named Asset Management Capital Company; Asset Management management activity and named partnerships; SEVF; Central/Eastern European funds/boards; successor structures; and the post-2012 family office. “Asset Management” is branding, not one continuous pool. Bancroft Johnson, pp. 35–43 and 78–92 · Stanford SEVF · AMC About Us [researcher inference · strong]

AMC reports more than 250 firm investments, while HBS variously reports more than 200 or roughly 250; these are shifting, self/institution-reported firm totals. SCOUT and DEALS route 37 attributable candidate families—Illumitronics, Behavioral Research Laboratories, Electroglas, Coherent, BioSurface Technologies, Chromatronix, Boole & Babbage, Qume, Tandem, Teradyne, Amgen, Applied Biosystems, Applied Micro Circuits, Hybritech, Biogen, IDEC, Conductus, Octel, Red Brick Systems, Remedy, Sierra Semiconductor, unidentified early AI, VisiCorp, Talkway, Verity, ViaCyte, Hypres, Reconstruction Capital II, Renaissance/Poland vehicles, a long-lived private/spinout pattern, the unnamed Microsoft competitor, the not-Amazon bookseller, Genentech, Intel, Apple, Southern Ute opportunities, and Stanford Engineering Venture Fund holdings. The two additions are distinct, target-specific funded failures established by the speaker-controlled 2005 panel turn. This is a discovery roster; 0/37 has a complete public target-person/vehicle-to-cash-flow chain. DEALS reconciliation [researcher inference · strong]

Chronology, attribution, and adverse controls

• D&J capital: Johnson and Draper separately remember $150,000 private capital split $75,000 each, while their leverage descriptions are not governing records. Draper, printed pp. 23–25 · Bancroft Johnson, printed pp. 20–27 [retrospective witness · moderate]
• AMC start: official/Bancroft sources use 1965 while HBS uses 1967; beginning to invest, branding a practice, legal formation, and first vehicle may be distinct events. Bancroft Johnson, printed pp. 35–40 · HBS award profile, p. 7 [researcher inference · provisional]
• Amgen: Johnson's board/capital role sits within a Bowes–Salser–Rathmann–Baddour–scientist–employee–syndicate story; “helped launch” does not prove sole founding, sourcing, or economic credit. Bancroft Johnson, printed pp. 45–73 · Caruthers, Amgen-formation paragraphs [researcher inference · strong]
• Tandem: Byers diligenced; Perkins/Kleiner Perkins formed and led the opportunity; Johnson supplied a small outside investment and Asset Management board representation. Perkins, printed pp. 38–43 · Byers, printed pp. 8–11 [retrospective witness · strong]
• Boole & Babbage: Johnson remembers an approximately $250,000 deal; Kolence reports founders invested $5,000 each and outside investors supplied $50,000 plus a $100,000 guarantee for 90%. The retrospective terms conflict and no cap table resolves them. Johnson Boole, printed pp. 4–5 · Kolence, printed pp. 18–20 [retrospective witness · provisional]
• Policy: the 1978 statute and 1979 regulation prove outcomes, not Johnson's individual contribution apart from NVCA, other investors, legislators, agencies, and broader pressure. Revenue Act of 1978 · 44 FR 37221 [researcher inference · strong]
• Adverse record: Johnson's speaker-controlled direct record supports funded VisiCorp and Talkway failures, an unidentified invested early-AI failure, a Microsoft-competitor loss, Intel/Genentech/Apple non-investments, shutdown reluctance, and long-lived weak cases. VisiCorp combines an ignored dependency warning with IBM-PC transition delay; Talkway combines a funded pivot with absent customer adoption and admitted demand misjudgment. Innovation video, VisiCorp 00:58:58–01:01:36 and Talkway 01:04:29–01:06:31 · CHM history, pp. 19–25 [investor-stated · moderate] The systemic critique identifies public-research, SBIC, tax, pension, inherited-capital, network, and era advantages; the bounded litigation sweep found no verified personal sanction, which is not proof none exists.

Historical, archival, and comparison coverage

The map covers postwar research/university infrastructure, SBIC leverage, tax/pension rules, early syndicates, later institutional fundraising, public markets, biotechnology, and Eastern European transition. Official law/regulation and a skeptical political-economy account keep public support, timing, team, access, and survivorship as rival explanations to personal skill. SBIC Act · Johnson 1998 Russia paper · systemic critique [researcher inference · strong]

Publicly accessible Bancroft, CHM, Stanford/OAC, HBS, Hoover, White House, company, SEC, GAO, and policy routes were processed through exact-name/title/date/company variants on 2026-08-20/21. Those legitimate routes did not expose D&J/AMC governing files; partnership agreements; investment schedules; LP letters; capital accounts; board minutes; private cap tables; complete cash-flow records; full MVS25/late-event media; regional-fund economics; or a complete outcome ledger. OAC M1939 finding aid · DEALS exhausted-route ledger [researcher inference · strong] No archive, company, or person was contacted.

The nearest completed comparators for later phases are Georges Doriot for public-corporation/permanent-capital and teacher influence; Laurance Rockefeller for family capital and research networks; Reid Dennis for personal/Group/corporate/LP vehicle evolution and shared panel/policy history; and Arthur Rock for people selection, syndication, Intel exclusion, governance, and finite partnerships. Bill Draper remains an unfinished but indispensable cofounder source, not an eligible completed-peer comparison.

Unresolved after exhausted public routes

1. The original 1982/1984 Entrepreneurial Climate, licensed Tanous chapter body, and Johnson course/case bodies did not surface in exact-title/author/library searches; only authenticated metadata, lawful preview, or later recap is usable. The 1998 Stanford-attributed paper is now identified and processed in full.
2. Every priority direct-voice family was processed to its public boundary; the 2002, 2005, and 2019 CHM videos have complete caption tracks with bounded Johnson turns, while MVS25 remains a host-selected fragment rather than a speaker-complete public program.
3. D&J and Asset Management entity/SBIC/partnership searches did not expose governing agreements, decision rights, or cash flows; the 1965/1967 discrepancy remains unresolved.
4. The 37-family ledger records the supported field and explicit shortfall for every family; no complete economic chain surfaced for any family.
5. Boole & Babbage, Amgen, Tandem, Electroglas, Applied Biosystems, Apple, and early AI were reconstructed only as far as the cited bodies allow; thin cases are not upgraded into full postmortems.
6. The complete 2005 caption track and direct Dennis-to-Johnson speaker boundary resolve VisiCorp and Talkway as distinct target-specific funded failures; company/date/vehicle/check/ownership/board/cash-flow fields remain limited to what Johnson's timestamped retrospective turn actually supplies.
7. SEVF, Johnson's 1998 paper, Renaissance history, and Reconstruction Capital II reports were processed; public bodies still do not expose vehicle-complete holdings, fees/carry, decision rights, or Johnson cash flows.
8. The final exact-name/company/title sweeps across SEC issuer filings, CHM, Stanford/HBS, SBA/GAO, OAC, CourtListener, and contemporaneous-press routes produced no new complete target-person cash-flow chain; the saturation boundary is therefore an evidenced public-record shortfall, not a claim that private archives contain nothing.

Source notes