Cancer Capital: It Sucks for Founders, Too!  🔗 **Source:** [Cancer Capital: It Sucks for Founders, Too!](https://www.anildash.com/2026/09/09/cancer-capital-founders/) > [!note]+ Document Notes > tags: [venture-capital, founder-investor-relations, collusion, antitrust, startup-funding, bootstrapping] > author: Anil Dash > source: anildash.com > reference: https://www.anildash.com/2026/09/09/cancer-capital-founders/ > date created: 2026-09-10 > > > [!info] Generated analysis — not the author's words. Sections below are machine-written > > from the captured source. Only text inside `> [!quote]` blocks is verbatim from the author. > > Do not quote or attribute any other line to the named author. > > # 🏷️ Title: Cancer Capital: It sucks for founders, too! > <!-- deep-read:sections start --> ## 🧭 Table of Contents - [📝 Summary](#-summary) - [📖 Main Content](#-main-content) - [🗒️ Notes](#️-notes) - [📚 References](#-references) - [🗣️ Paraphrase](#️-paraphrase) - [📰 Newsletter](#-newsletter) - [✅ Action Items](#-action-items) - [🧪 Hypotheses](#-hypotheses) - [📊 Methodology](#-methodology) - [❓ Questions](#-questions) - [🔮 Future Research](#-future-research) - [💡 Implications](#-implications) - [🧠 Learning Styles Assessment](#-learning-styles-assessment) - [📊 Accessibility Assessment](#-accessibility-assessment) - [🌍 DEIB Assessment](#-deib-assessment) - [✅ Logical Validity](#-logical-validity) - [🛡️ DARVO Analysis](#-darvo-analysis) - [🎯 Conclusion](#-conclusion) - [🔄 Refactor Appendix](#-refactor-appendix--cancer-capital-it-sucks-for-founders-too) ### 🔗 References - [📖 Main Content — the collusion mechanism and the "Any way out" remedies](#-main-content) - [🗒️ Notes — section-by-section walkthrough](#️-notes) - [🔄 Refactor Appendix — atoms and permanent candidates](#-refactor-appendix--cancer-capital-it-sucks-for-founders-too) ## 📝 Summary - **What happened.** Anil Dash argues, as the third installment in his "Cancer Capital" series, that a small group of hyper-scale VC firms now routinely collude against the founders pitching them: firms refuse to sign NDAs, then trade confidential pitch details with each other before any of them commits, giving the biggest firms near-complete visibility into an entire market while founders get worse deal terms. - **Why it matters.** The piece names a specific, previously informal mechanism — no-NDA norms plus the collapse of the "don't fund competitors" rule — and argues it has flipped from a minor inconvenience into a structural information monopoly once a fund is large enough that most founders in a category have to pitch it. - **Key stakeholders.** Founders raising past roughly $100M in valuation; the "Cancer Capital" mega-funds Dash doesn't name individually in this piece (established in his prior installments, linked but not reproduced here); the smaller NYC-era firms he credits for holding older founder-friendly norms (Union Square Ventures, Bloomberg Beta); limited partners such as public pension funds; state regulators, since he says federal antitrust enforcement is currently unavailable. - **Risks and tradeoffs.** The piece's most serious claim — that a firm can pass on a founder's idea, then fund a clone through its own network to kill the original company — is sourced only to unnamed founders who "won't go on the record," which the Fallacies section below flags as unfalsifiable as written. The six "way out" remedies also carry real costs Dash names but doesn't fully weigh (bootstrapping foregoes capital, avoiding mega-funds forecloses the largest checks, follow-on rounds can force contact with the same firms anyway). - **Bottom line.** Dash distinguishes ordinary co-investment syndication (which happens after a firm commits, with the founder's knowledge) from what he's describing (firms comparing confidential notes before any of them commit) — a real, checkable conceptual distinction, not just a restated grievance. The argument that this behavior scales with fund size, and gets structurally worse as more capital concentrates in fewer firms, holds together; the anonymous clone-and-kill allegation is the piece's weakest evidentiary link. ## 📖 Main Content Dash opens by locating this piece inside a series: two earlier posts covered how VC evolved over roughly two decades to get worse, and this one moves from that framing to concrete effects on founders — "people who have a great idea for a product or a service" — as distinct from a newer cohort he calls people who "identified a pile of money that they wanted to grab" and will "suck up to whatever investors" to get it. He sets that second group aside for later in the piece. **How the old deal worked.** Founders bootstrapped as far as they could on their own or with friends-and-family money, then treated outside investors as a "necessary evil," per the old "vulture capital" framing Dash says was conventional wisdom. On the investor side, he credits firms like Union Square Ventures (New York) for competing on founder-friendliness: a legible, publicly understandable investment thesis that let founders self-select out before wasting months on a doomed pitch, and — critically — a norm that a VC would not invest in a company competing with one already in its portfolio, because sitting on two competing boards is an obvious, costly conflict. He cites his own 2013 pitch to Bloomberg Beta, which had published its entire operating manual on GitHub and let his company publish its term sheet too, as an example of firms competing on transparency rather than leverage. **The no-NDA norm and how it curdled.** For decades, VCs have publicly refused to sign NDAs — Dash names Brad Feld, Mark Suster, and Fred Wilson as having blogged this position years ago, back when it was a reasonable position because firms genuinely wouldn't fund a founder's direct competitor. He argues the same no-NDA policy is now "sinister" in effect: founders still hand over a full financial model, customer pipeline, product roadmap, and competitive self-assessment, to firms who refused confidentiality *before the founder ever walked in the room* — a one-sided disclosure obligation dressed up as an old, neutral industry norm. **The collusion mechanism itself.** Dash's central claim, stated twice for emphasis, is that "major venture capital firms now routinely collude against founders, which means those founders end up with worse terms for their deals." His evidence is a recurring pattern he says he's personally observed across many founders' pitch processes: a partner at one firm demonstrates unprompted familiarity with a startup's details before the founder has covered them, sometimes naming another firm outright ("I was talking to [X] over at [other venture firm], and he thinks this is really interesting") or framing the disclosure as a favor to the founder ("We were thinking we might team up with that other firm and we'll go in together on your next round"). He draws a specific, load-bearing distinction here: legitimate syndication happens *after* a firm has committed, with the founder's knowledge and consent; what he's describing is firms — some of whom will pass entirely — comparing confidential notes about a company *before* any commitment exists. He also notes this predates today's AI-assisted deal analysis tools, i.e., it isn't a new problem created by better data tooling. He points to one piece of real enforcement precedent for the underlying conflict-of-interest problem: the DOJ, in 2023, "quietly pushed a dozen directors off nine company boards" over illegal interlocking directorates, mostly at private equity firms — but he's explicit that this only reached the board level, not the pitch-meeting level where he says the actual information-sharing happens unchecked ("Nobody's looking at the pitch meeting"). **How "Cancer Capital" scale makes it worse.** Dash argues three compounding effects specific to the largest funds. First, a founder essentially has to pitch the small handful of mega-funds by the time a company nears a $100M valuation, or their absence from the cap table reads as a negative signal — so every competitor in a hot category ends up disclosing to the same few firms, none of whom signed anything. Second, the old anti-competitor-investment norm is simply gone: firms now openly invest in multiple, even several, directly competing companies in a category, describing it as "taking a position on the category" rather than backing one company. Third, once a mega-fund also holds board seats or information rights in several competitors, it has access to monthly numbers, real churn, staffing, and compensation data no regulator or outside analyst sees — Dash's phrase for this position is that "the extractive VCs aren't just investing in the market — they're the only party that can see it." **The clone-and-kill allegation.** The most serious specific claim in the piece: if a Cancer Capital firm likes an idea but not the founder's ethics — his parenthetical gloss is "the founder *has* morals" — it can, per Dash, "elect to merely choose someone in their network to create a clone of the idea, and then hyper-fund that clone in order to kill the company that just pitched them." Dash frames this as something he's heard about firsthand from multiple founders, none willing to go on record, and says the unnamed evidence ("receipts") he's personally seen makes him "100% convinced." No named instance, document, or corroborating third party appears in the piece — this is Dash's allegation, presented as more than speculation but without checkable support in the text itself. **Selection effects on who becomes a founder.** Dash argues the current system doesn't just tolerate the "pile of money" opportunists he set aside in the opening — it actively selects for them: given a choice between a founder who will "fight you on terms" and one who will "execute the ideas laid out in the manifesto on your fund's homepage," a Cancer Capital firm rationally prefers the latter. He also draws an explicit analogy to social media network effects — toxic platforms retain users because "everybody else is there" — and notes, without naming individuals, that the people running Cancer Capital funds built much of their wealth from "the biggest, worst social networking platforms." **Six ways out.** Dash closes with six remedies, the first four aimed at avoiding the trap and the last two at pressuring the system itself, explicitly caveated as none being sufficient alone: (1) bootstrap and avoid VC where possible, citing Microsoft and Apple reaching early milestones without it; (2) build leaner using current tools, including AI, and lean on open-source communities instead of capital; (3) "leak-proof the deck" — assume everything you disclose reaches a competitor, sequence what you share, or go fully transparent so there's nothing to leak; (4) work with the "99%" of more conventional VC firms, while cautioning that a later follow-on round can still force contact with a Cancer Capital fund; (5) push for regulation, which he says is currently blocked federally by what he calls "corruption under an authoritarian regime," leaving state and local action as the more realistic near-term lever; (6) pressure limited partners — public pension funds with existing responsible-investment commitments — to divest from Cancer Capital funds. ## 🗒️ Notes - **Series placement:** this is post 3 in Dash's "Cancer Capital" series — post 1 (2026-09-02) introduces the term; post 2 (2026-09-04) covers VC's historical evolution; this post applies both to founders specifically. A companion piece from 2025-10-24, "Founders over Funders," is also linked as prior framing for the founder/opportunist distinction drawn in this piece's opening. - **Named old-guard evidence:** Union Square Ventures (NYC) credited for a legible thesis + founder-friendly reputation; Bloomberg Beta's 2013 public GitHub operating manual and term-sheet transparency, from Dash's own pitch experience — this is first-person, direct evidence, not hearsay, and is the strongest-sourced claim in the piece. - **NDA-norm sourcing:** Brad Feld, Mark Suster, and Fred Wilson each publicly blogged the no-NDA VC norm "many years ago" — Dash treats this as long-standing, well-documented industry practice, not something he's asserting alone. - **The $100M valuation threshold** is presented as an observed pattern ("if it's not in the initial round... then certainly by the time...") rather than a cited statistic — no survey, dataset, or source is given for this figure; treat it as Dash's practitioner estimate, not a measured benchmark. - **DOJ interlocking-directorate action (2023):** cited via a Cooley law-firm write-up, "a dozen directors off nine company boards," mostly at private equity firms. This is real, checkable enforcement — but Dash is explicit that it addressed board-level conflicts, not the pitch-meeting information-sharing he's describing; he does not claim regulators have acted on the latter. - **Information rights specifics:** monthly numbers, real churn, staffing plans, and compensation are named as the category of data mega-funds get via board seats/information rights that goes beyond what's in a normal pitch deck — this is the mechanism behind the "X-ray vision" claim, not just a rhetorical flourish. - **The clone-and-kill claim is explicitly unsourced beyond Dash's own say-so** — "nobody will go on the record" and "the receipts I've seen make me 100% convinced" are both his framing; no external corroboration exists in the piece. Distinguish this clearly from the collusion-in-pitch-meetings claim, which Dash frames as something he's personally witnessed and reports concretely (the "I was talking to X" and "we might team up" lines read as near-verbatim recollections, not composites). - **"Fake founders" framing:** Dash's opening parenthetical about opportunist founders ("identified a pile of money... willing to suck up to whatever investors") is explicitly revisited later as a *product* of Cancer Capital selection dynamics, not an unrelated aside — the piece's structure deliberately closes that loop. - **Six remedies, explicitly incomplete:** Dash frames the list as "no individual action can solve a systemic problem, but all of these tactics together can begin to change these systems" — he does not claim any single remedy is sufficient, including the regulatory and LP-pressure options he clearly favors. - **Regulatory claim carries a specific political assertion:** "corruption under an authoritarian regime" is Dash's characterization of why federal action is currently unavailable — stated as fact in the piece, not hedged as opinion; readers should treat this as Dash's political position, not a neutral premise. ## 📚 References - [Cancer Capital (part 1)](https://www.anildash.com/2026/09/02/cancer-capital/) - [VC Didn't Use to Work Like This (part 2)](https://www.anildash.com/2026/09/04/vc-didnt-use-to-work-like-this/) - [Founders over Funders](https://www.anildash.com/2025/10/24/founders-over-funders/) - [The VC Three (2006)](https://www.anildash.com/2006/07/14/the_vc_three/) - [Bloomberg Beta's public operating manual (GitHub)](https://github.com/Bloomberg-Beta/Manual) - [Ten Tips Guaranteed to Improve Your Startup Success](https://www.anildash.com/2013/03/28/ten_tips_guaranteed_to_improve_your_startup_success/) - [VC QAnon](https://www.anildash.com/2023/07/07/vc-qanon/) - [Brad Feld — Why Most VCs Don't Sign NDAs](https://feld.com/archives/2006/02/why-most-vcs-dont-sign-ndas/) - [Mark Suster — On NDAs and Confidentiality](https://medium.com/both-sides-of-the-table/on-ndas-and-confidentiality-dfa741ef58cb) - [Fred Wilson — Why VCs Don't Sign NDAs](https://avc.com/2006/02/why_vcs_dont_si/) - [Cooley — Assessing Overlapping Boards After DOJ Crackdown (2023)](https://www.cooley.com/news/insight/2023/2023-05-25-assessing-overlapping-boards-after-doj-crackdown) - [TechCrunch, Venture Capital, Record Labels, and Getting What You Asked For](https://www.anildash.com/2009/09/25/techcrunch_venture_capital_record_labels_and_getting_what_you_asked_for/) - [Funding a Startup Without VC](https://www.anildash.com/2011/04/18/funding_a_startup_without_vc/) ## 🗣️ Paraphrase Dash's argument is that the informal decency norms that used to govern venture capital — a legible thesis, no funding a portfolio company's direct competitor, and a founder-friendly posture firms competed to demonstrate — have broken down at the top of the market, and the mechanism by which they broke down is specific rather than vague: VCs' decades-old refusal to sign NDAs, once a minor inconvenience because firms genuinely avoided competitive conflicts, now means founders disclose everything to firms who share what they learn with each other before any of them commits. He's careful to separate this from ordinary syndication, which happens with a founder's knowledge after a firm has already committed capital. The scale problem is what turns this from an occasional bad actor into a structural one. Once a handful of mega-funds are functionally mandatory to pitch by the time a startup is worth real money, and once those same funds routinely back several competitors in one category while holding information rights that go beyond the pitch deck, they end up with more visibility into an entire market than regulators or the companies in it. His most serious allegation — that a firm can quietly fund a clone of a rejected idea to kill the company that pitched it — is the one claim in the piece resting entirely on unnamed sources, which he acknowledges but doesn't resolve. The piece ends practically, not just diagnostically: six concrete moves, from bootstrapping to leak-proofing a pitch deck to pressuring limited partners, offered as partial mitigations rather than a full fix. ## 📰 Newsletter Anil Dash's third "Cancer Capital" post is worth a read for anyone raising money or advising someone who is — it names a mechanic that's easy to sense but hard to prove: mega-funds trading notes on your pitch before any of them have committed to it, using the industry's old "we don't sign NDAs" norm as cover. The most useful part isn't the outrage, it's the distinction he draws between normal co-investment syndication (after commitment, with your knowledge) and what he's describing (before commitment, without it) — a line worth knowing if you're deciding what to put in a pitch deck versus hold back for a signed term sheet. His six-item "way out" list, especially "leak-proof the deck" and leaning on open-source/community-supported building to stay capital-light, applies directly to indie and bootstrapped product builders who'd rather not need a mega-fund's check at all. ## ✅ Action Items #### DevRel - Not applicable to this source — the piece is about VC-founder power dynamics, not developer advocacy, docs, or community programs. #### OSPO - One partial connection: Dash names "open source and open communities" as a concrete way to build "more efficiently" and reduce dependence on outside capital (remedy #2). Worth flagging to any OSPO evaluating why a project chose open-source distribution over a funded roadmap — this piece is a founder-side argument for that choice, not an OSPO-side one. - Otherwise not applicable — no licensing, SBOM, or governance content. #### Community - The same open-source/community point applies here: Dash frames "remaining closely connected to a community that can support you" as a direct substitute for VC dependency. That's a community-management-adjacent claim (community as capital substitute) worth citing if building the case for investing in community infrastructure, but the piece doesn't go further than the one sentence. - Otherwise not applicable — no maintainer, contributor, or sponsor content. #### WordPress - Not applicable to this source across all subteams (Core, Core Performance, Core AI, Core Program Team, Mobile, Playground, Community, Training, Marketing, Support, Design, Documentation, Photos, TV, Polyglots, Accessibility, Meta, CLI, Hosting, Tide, Openverse, Themes, Plugins) — WordPress is never mentioned; this is a general startup-funding piece. #### The WPOCC - Not applicable directly — no fellowship, funding, or governance content specific to The WPOCC. The LP-divestment remedy (pressuring funding sources with responsible-investment commitments to withdraw support from bad actors) is structurally the same lever The WPOCC's funding-transparency work uses; worth noting as a parallel mechanism, not a direct action item. #### FAIR - Not applicable — no federation, provenance, or governance-artifact content in this source. ## 🧪 Hypotheses - H1: Startups that pitch a "Cancer Capital"-tier fund receive measurably worse deal terms (valuation, liquidation preference, board control) than comparable startups that raise exclusively from the "99%" of conventional VC firms. - H2: Founders who withhold specific categories of information (full financial model, detailed competitive self-assessment) until after a term sheet is signed close rounds on terms no worse, on average, than founders who disclose everything upfront. - H3: VC firms that hold board seats or information rights in two or more directly competing companies in the same category can be identified from public filings/portfolio pages, and their portfolio companies in that category show correlated strategic moves (pricing, hiring, product timing) at a rate higher than chance. - H4: The rate at which founders report a firm "revealing familiarity" with their business before disclosure (Dash's anecdotal pattern) is higher for mega-funds than for firms outside the top-AUM tier. - H5: Public pension funds and other LPs with stated responsible-investment policies, when presented with documented collusion patterns, measurably reduce or condition future commitments to the named funds within 1–2 fund cycles. - H6: The "clone and kill" allegation, if true even in a minority of cases, would leave a detectable signature — a fund passing on a company, then funding a near-identical company in the same category within a short window, run by a founder connected to the fund's existing network. ### 📊 Methodology - H1: Compare term sheets (via founder surveys, e.g., through YC, Carta's State of Startups data, or a purpose-built anonymous survey) across cohorts defined by lead-investor AUM tier; control for sector, round size, and founder track record. - H2: A/B or matched-cohort study of disclosure timing across founders raising in the same period and sector, using self-reported or law-firm-aggregated deal-term data as the outcome. - H3: Scrape portfolio pages and Form D/board-composition filings where available; build a co-investment/competitive-overlap graph; test for correlated strategic timing using public product-launch and hiring data (LinkedIn, job postings) as a proxy. - H4: Structured founder survey (pre-registered categories: "firm showed unprompted familiarity," "firm named another firm," "firm offered to 'team up'") segmented by lead-firm AUM. - H5: Track LP commitment announcements and fund-of-funds allocations to named large VC firms over time, cross-referenced against public reporting of collusion or antitrust allegations; interrupted time-series analysis around major press events. - H6: Build a dataset of (a) publicly reported "passed on X, funded similar company Y" cases and (b) the network ties between fund and founder in case Y; test whether the rate of such pairs exceeds a null model of independent convergent invention in the same category. ### ❓ Questions - Does "collusion" in Dash's sense (informal note-comparing before commitment) meet any existing legal definition — antitrust, fiduciary, or otherwise — or is it presently a reputational-only harm with no cause of action? - How would a founder or a researcher distinguish the clone-and-kill pattern from ordinary convergent invention, where multiple teams independently pursue an obviously good idea in the same window? - Is the DOJ's 2023 interlocking-directorate enforcement (board-level) extensible to pitch-meeting-level information sharing under current law, or does it require new statutory authority? - What share of the "99%" of conventional VC firms Dash credits with holding the line actually avoid follow-on rounds that route through a Cancer Capital fund, versus merely avoiding leading such rounds? - Does the $100M-valuation threshold for "mandatory" mega-fund pitching hold across sectors (e.g., deep tech vs. consumer software), or is it specific to the categories Dash has personally observed? ### 🔮 Future Research - No dataset in the piece ties fund size/AUM to measured deal-term outcomes — this is the single most valuable gap, since it would convert Dash's practitioner pattern-recognition into a checkable claim. - No legal analysis of whether the described conduct (comparing notes pre-commitment, absent a signed NDA) exposes firms to existing antitrust or tortious-interference liability, versus needing new regulation as Dash assumes. - No named or documented instance of the clone-and-kill allegation — a single verifiable case, even anonymized with corroborating detail, would substantially change the evidentiary status of the piece's strongest claim. - No tracking of whether any LP has actually divested from a named Cancer Capital fund over these concerns, which would test remedy #6's real-world traction rather than its theoretical plausibility. ### 💡 Implications - **Practice.** Founders can act on the syndication-vs-collusion distinction immediately: treat any pre-commitment familiarity with your business from an uninvolved firm as a signal to ask directly who they spoke to, and sequence disclosure (withhold granular financials/competitive detail until a term sheet exists). - **Policy/governance.** Dash's own read is that federal antitrust enforcement is currently unavailable; the piece implies state-level unfair-competition or anti-collusion statutes are the more realistic near-term lever, and that the DOJ's board-level 2023 action is a precedent for a lower-level version of the same theory, not proof it already applies. - **Security.** Information rights (monthly numbers, churn, staffing, compensation) constitute a real data-exposure surface once a firm invests across multiple competitors in a category — a founder due-diligence checklist could reasonably ask a prospective investor to name every other company they hold information rights in within the same category. - **Education.** The no-NDA norm's history (Feld/Suster/Wilson posts) is decades old and no longer taught with the caveat Dash adds here — that it depended on a now-absent no-compete-investment norm to be safe for founders. That's a concrete, teachable correction for founder-education content. - **Funding models.** The piece's own remedies (bootstrapping, open-source/community-supported building, LP pressure) function as a funding-model critique in practice — each one is a way to either avoid VC entirely or make the VC relationship itself more accountable, rather than a call to reform VC internally. ### 🧠 Learning Styles Assessment - **Reading level estimate:** Flesch-Kincaid Grade Level ≈ 12.5 (computed from the captured text: ~3,088 words, 117 sentences, ~4,659 syllables). Basis: long compound-complex sentences with frequent parenthetical asides, insider terminology used without definition (term sheet, information rights, syndication, LP, cap table), and multi-clause argumentative sentences rather than short declaratives. #### 👀 Visual learners - **Present:** one header image; section breaks marked with descriptive H2 subheadings ("What about the investors?", "Stacking the deck," "Cancer Capital colludes against you," "Any way out?"). - **Improve:** a simple before/after diagram (old norms vs. current practice: NDA policy, competitor-investment rule, disclosure obligations) would make the piece's core mechanism scannable without reading the full argument. #### 🔊 Aural learners - **Present:** none — text-only blog post with no audio or narrated version. - **Improve:** a short audio read-through or podcast adaptation would suit this argumentative, narrative-driven piece well, since it reads close to spoken argument already. #### 🗣️ Verbal learners - **Present:** clear definitional work distinguishing syndication from collusion, and "how-to-do" concrete examples (the "I was talking to X" and "we might team up" dialogue snippets ground the abstract claim in recognizable scenarios). - **Improve:** a short glossary of VC terms used without definition (information rights, LP, syndication, term sheet) would help readers outside the startup-funding world follow the mechanism. #### 🧩 Kinesthetic learners - **Present:** the six "way out" remedies are actionable and list-formatted, closest thing in the piece to a checklist. - **Improve:** a due-diligence checklist a founder could literally bring into a pitch meeting (e.g., "ask which competitors in this category the firm holds information rights in") would convert the argument into a reusable tool. #### 🧠 Logical learners - **Present:** the explicit conceptual distinction between syndication and collusion, and the three-part breakdown of what changed at "Cancer Capital" scale (mandatory pitching, no competitor rule, compounding information rights). - **Improve:** a simple causal chain diagram (no-NDA norm → competitor-rule collapse → mega-fund scale → market-wide information monopoly) would make the argument's logical structure explicit rather than narrative. #### 🧑‍🤝‍🧑 Social learners - **Present:** the piece explicitly invites comparison ("always, always compare notes with other founders") as a defensive tactic. - **Improve:** a structured, anonymized founder reporting mechanism (referenced only as an idea, not built) would turn the "compare notes" advice into an actual social/collective tool rather than informal advice. #### 🧘‍♀️ Solitary learners - **Present:** the six remedies are individually actionable by a single founder without needing collective action (bootstrap, build lean, leak-proof the deck). - **Improve:** none needed — the piece already supports self-directed action well for a solo founder. #### 🌿 Naturalistic learners - **Present:** the "Cancer Capital" metaphor itself is a biological/systemic framing (a growth that damages the host system), used consistently across the series. - **Improve:** not a strong fit otherwise; no ecological or lifecycle content beyond the framing metaphor. ### 📊 Accessibility Assessment *Assessed from the captured HTML only — visual contrast and live rendering were not directly reviewed.* #### Screen reader semantics - **Present:** proper heading hierarchy (H2 subsections: "What about the investors?", "After the Good Old Days," "Stacking the deck," "Cancer Capital colludes against you," "Any way out?"); the "Any way out?" remedies use a real semantic `<ul>/<li>` list rather than visual-only formatting. - **Improve:** the header `<figure>` image's `alt` text duplicates the post title rather than describing the image itself — WCAG 2.2 SC 1.1.1 (Non-text Content) is technically met but the alt text carries no independent information for a screen-reader user. #### Keyboard navigation - **Not observed** — no interactive elements (forms, embedded widgets) in the captured content; standard link/paragraph structure implies no custom focus traps, but this wasn't verified against the live page (WCAG 2.2 SC 2.1.1, 2.4.3). #### Contrast and readability - **Not observed** — text/background contrast wasn't visually verified against the live site (WCAG 2.2 SC 1.4.3, 1.4.11); flag for direct review before treating as compliant. #### Text resizing and reflow - **Not observed** — responsive behavior at 200% zoom wasn't tested against the live page (WCAG 2.2 SC 1.4.4, 1.4.10). #### Media and captions - **Not applicable** — no audio or video content in this post. #### Forms and interactions - **Not applicable** — no forms or interactive input in the captured content. #### Standards reference - Gaps above map to WCAG 2.2 AA: 1.1.1 (partial — alt text present but low-value), 1.4.3, 1.4.4, 1.4.10, 1.4.11, 2.1.1, 2.4.3. Most items require direct review of the live rendered page to confirm or close. ### 🌍 DEIB Assessment #### Cultural breadth - **Present:** the piece's positive examples are geographically specific (Union Square Ventures and Bloomberg Beta, both New York-based), which Dash uses deliberately to credit a specific regional tech community. - **Improve:** the analysis is otherwise US-centric throughout (DOJ enforcement, US pension-fund LPs, "authoritarian regime" framed as a US-specific political claim); no non-US VC ecosystem or founder experience is discussed, which may limit how directly international readers can apply the remedies. #### Inclusive language - **Present:** consistent, direct second- and third-person address; no gendered generic terms for "founder" or "investor." - **Improve:** the piece uses politically loaded terms ("extremist political positions," "authoritarian regime") presented as established fact rather than flagged as Dash's own position — readers should be told explicitly (as this note does) that these are his characterizations, not neutral premises, before treating the argument's political framing as settled. #### Neurodiversity support - **Present:** clear H2 section breaks with descriptive (non-cute) headers make the argument's structure trackable across a long piece. - **Improve:** several sections run as single long paragraphs with multiple embedded sub-claims (e.g., the paragraph introducing the pitch-meeting collusion pattern); breaking these into shorter paragraphs or a summary callout per section would aid readers who rely on visual chunking. #### Representation - **Present:** the piece explicitly notes that VC firms once courted "underrepresented founders" as a stated value (2013-era inclusivity marketing) — Dash treats this history evenhandedly rather than dismissing it. - **Improve:** every named individual in the piece (Brad Feld, Mark Suster, Fred Wilson) is male, and no female or non-binary investor or founder is named or quoted as a source — the "underrepresented founders" reference is the only place representation beyond this default appears, and it's mentioned in passing rather than developed. #### Psychological safety - **Present:** the piece names founders' real distress around pitch meetings ("break into tears," "wake up with panic attacks") without minimizing it, and frames founder anxiety as a reasonable response to real power imbalance rather than a personal failing. - **Improve:** none needed for this content — no sensitive material requiring an additional content warning. ### ✅ Logical Validity #### Strong reasoning - **Conceptual distinction, not just assertion (syndication vs. collusion):** Dash doesn't just call the behavior "collusion" and move on — he explicitly defines what legitimate co-investment (syndication) looks like (after commitment, with founder knowledge) and contrasts it against the pattern he's describing (before commitment, without it). This is a real, checkable distinction that strengthens the central claim rather than leaving it as an undifferentiated grievance. - **Mechanistic reasoning on information asymmetry:** the claim that a firm holding board seats/information rights across several competitors in a category ends up with more market visibility than regulators is a specific causal chain (information rights → cross-competitor data → aggregate market visibility), not a vague assertion of unfairness — it follows from named mechanisms (churn, staffing, compensation data) rather than just the fact of large fund size. #### Fallacies - **Unfalsifiable anecdotal claim presented with high confidence — the clone-and-kill allegation.** Location: "I've heard about it firsthand, though nobody will go on the record... The receipts I've seen make me 100% convinced." Why it's weak: no named instance, document, date, or corroborating party is given; "100% convinced" is asserted certainty resting entirely on evidence the reader cannot evaluate. Fix: even one specific, anonymized-but-detailed case (timeline, category, general fund tier) would let readers assess plausibility instead of taking the claim on Dash's authority alone. - **Unsupported statistic presented as observed pattern.** Location: "by the time a company has reached a valuation of about $100M or so, they will have been expected to pitch one of the small handful of Cancer Capital funds." Why it's weak: no source, survey, or dataset backs this threshold — it reads as a specific number but functions as a rounded impression from Dash's own experience. Fix: label it explicitly as a practitioner estimate, or cite even informal survey data (e.g., from founder communities) if it exists. - **Presented-as-fact political characterization.** Location: "due to the level of corruption under an authoritarian regime" (given as the reason federal regulation is currently unavailable). Why it's weak: this is a contested political claim stated with the same declarative confidence as the piece's checkable industry claims, which risks readers extending unwarranted confidence to the political premise because it's phrased identically to the sourced material around it. Fix: flag explicitly as the author's political position, separate from the VC-industry claims that carry named sourcing. #### Tone of voice - Direct, first-person, argumentative and unapologetically opinionated; frequent italics and parentheticals for editorial asides ("(spoilers: it got worse!)," "(i.e. the founder *has* morals)"); confident, undedged declarative sentences even where evidence is thin (the clone-and-kill claim); empathetic rather than clinical when describing founders' experience of pitch-meeting stress. ### 🛡️ DARVO Analysis - **Not applicable in the standard sense.** DARVO (Deny, Attack, Reverse Victim and Offender) describes a rhetorical move made by an *accused party* responding to an allegation. No accused party appears in this text to respond — Dash is the one making the allegations (collusion, and separately, clone-funding), and no VC firm's rebuttal, denial, or counter-framing is present in the captured content. Applying DARVO to Dash himself would mischaracterize what DARVO measures: he is alleging harm done to founders, not responding to an accusation made against him. - **The actual rhetoric worth naming instead:** the series' branding choice, "Cancer Capital," is a deliberately loaded, non-neutral label applied to an entire class of firms — a value judgment embedded in the term itself rather than argued for fresh in each post (the term's justification lives in the earlier installments, linked but not reproduced here). That's a legitimate rhetorical strategy for an opinion series, but it means a reader encountering this post in isolation is handed the conclusion (these firms are a "cancer") before the argument for it is made in this specific piece. - **No victim-blaming or victim/offender reversal appears.** Founders are consistently framed as the party being acted upon, not as complicit or responsible for what's done to them; the one group Dash treats critically among founders — the "pile of money" opportunists — is explicitly separated from the founders he's defending, not conflated with them. ## 🎯 Conclusion Dash's core claim is specific enough to be checkable: the mechanism (no-NDA norm plus the collapse of the anti-competitor-investment rule, at mega-fund scale) is a real, describable shift from the founder-friendly norms he documents from a dozen-plus years ago, not a vague complaint about VCs being difficult. The syndication-vs-collusion distinction is the piece's strongest analytical move; the clone-and-kill allegation is its weakest evidentiary one, resting entirely on unnamed sources. Practical next step for a founder reading this: before your next pitch, ask directly which other companies in your category the firm holds a board seat or information rights in, and treat any unprompted familiarity with your business from an uninvolved partner as a reason to ask, specifically, who they've spoken to — the test this piece itself recommends but never shows anyone actually running. ## 🔄 Refactor Appendix — Cancer Capital: It sucks for founders, too! > Density: 5 Literature atoms (H4), 2 Permanent candidates (H4). > Konik claim-header rules: present tense, declarative, ≤10 words, searchable, unique. > Atomicity: one declarative sentence per atom. ### 📚 Literature Atoms (H4) #### [L1] Mega VC firms now routinely collude against founders > [!quote] > major venture capital firms now routinely collude against founders, which means those founders end up with worse terms for their deals. **Context & use**: - Stated twice in the piece for emphasis, directly after Dash previews it as something that will "sound shocking." - Load-bearing thesis for the entire piece; every subsequent mechanism (no-NDA norm, competitor-investment rule collapse, information rights) is offered as support for this claim. **Tags**: #xref #aphorisms <!-- atomicity: 1 idea ✓ --> #### [L2] Firms reveal they compared notes on your pitch > [!quote] > We were thinking we might team up with that other firm and we'll go in together on your next round. **Context & use**: - Presented as a real, recalled quote from a VC partner, framed by the firm as a favor to the founder. - Dash's evidence that firms disclose cross-firm conversations about a still-uncommitted company casually, not as a rare slip. **Tags**: #xref <!-- atomicity: 1 idea ✓ --> #### [L3] Alleged: firms fund clones to kill rejected founders > [!quote] > create a clone of the idea, and then hyper-fund that clone in order to kill the company that just pitched them. **Context & use**: - The piece's most serious and least-corroborated claim — sourced to unnamed founders, no documented instance given. - Directly relevant to the Fallacies finding above; treat as an alleged practice, not an established fact. **Tags**: #xref <!-- atomicity: 1 idea ✓ --> #### [L4] Cross-competitor info rights give VCs sole market visibility > [!quote] > the extractive VCs aren't just investing in the market — they're the only party that can see it. **Context & use**: - Follows directly from naming the specific data (churn, staffing, compensation) that information rights expose beyond a pitch deck. - Grounds the "X-ray vision" framing in a concrete mechanism rather than metaphor alone. **Tags**: #xref #aphorisms <!-- atomicity: 1 idea ✓ --> #### [L5] Bootstrap: live within your means, don't raise VC > [!quote] > Live within your means, don't raise VC. **Context & use**: - First and most direct of the six "way out" remedies; supported by naming Microsoft and Apple as pre-VC-milestone examples. - A blunt, aphoristic distillation of the piece's most actionable practical advice. **Tags**: #xref #aphorisms <!-- atomicity: 1 idea ✓ --> ### 🧠 Permanent Candidates (H4) #### [P1] Ask what competitors a VC already has information rights in **Claim** (Courtney's voice): Before I disclose anything sensitive to a prospective investor, I should ask directly which other companies in my category they hold a board seat or information rights in, since that's the actual exposure — not whether they'll sign an NDA they were always going to refuse. **Evidence pointer**: [[#L4]] **Scope & conditions**: Most relevant once a fund is large enough to hold multiple competing portfolio companies in one category; less relevant for early, thesis-stage checks from small funds without competing category bets yet. **Links** (≥1 with annotation): - [[Due diligence reciprocity]] — because extends: due diligence isn't just the investor vetting the founder, it runs both directions. - [[Information asymmetry]] — because applies-in: any negotiation where one party has structurally more visibility than the other benefits from naming the asymmetry explicitly before disclosing further. <!-- atomicity: 1 idea ✓ --> #### [P2] Distinguish alleged harm from allegations I can't verify **Claim** (Courtney's voice): When a source makes a serious, plausible allegation that rests entirely on unnamed sources, I should treat the underlying pattern as worth tracking and the specific allegation as unverified — not dismiss the pattern because one claim in the piece is thinly sourced. **Evidence pointer**: [[#L3]] **Scope & conditions**: Applies specifically to allegations of active harm (fraud, retaliation, sabotage) as opposed to documented structural claims (norms, incentives, public precedent), which can be evaluated on their own evidentiary footing within the same source. **Links** (≥1 with annotation): - [[Source evaluation]] — because applies-in: any deep-read source that mixes well-sourced structural argument with a thinly-sourced serious allegation. <!-- atomicity: 1 idea ✓ --> ### ✏️ Fleeting Queue - [F1] Check whether any public dataset (Carta State of Startups, YC founder surveys) ties lead-investor fund size to deal-term outcomes — would directly test H1. — #xref - [F2] Watch for whether any LP publicly discloses divesting from a named "Cancer Capital" fund over these concerns — would test remedy #6's real-world traction. — #xref ### → Things 3 (max 2 actions) - [ ] Read Dash's parts 1 and 2 of the Cancer Capital series (2026-09-02, 2026-09-04) to confirm which specific firms he's characterizing before citing this piece as a source in anything published — context #verification #things - [ ] Draft a founder due-diligence question ("which competitors in my category do you hold information rights in?") for reuse in any future pitch-prep or founder-advising notes — context #pkm-extraction #things <!-- deep-read:sections end --> 📄 **Full text:** [[Full Document Contents/Articles/Cancer Capital It Sucks for Founders, Too!.md|Cancer Capital: It Sucks for Founders, Too! by Anil Dash]] ## Highlights - major venture capital firms now routinely collude against founders, which means those founders end up with worse terms ([View Highlight](https://read.readwise.io/read/01m25w10rzad233khs4rckf8y8)) - Note: Dash's load-bearing thesis. Sourced only to founders who won't go on record — alleged, not documented. - Tags: venture-capital, alleged