# Investor thesis fit: prove the match, not the keyword Turn a broad investor list into a defensible research shortlist by comparing the exact business you are building with the mandate a fund can actually support. Source: https://12inf.ai/resources/find-current-thesis-fit Updated: 2026-10-14 ## Start with a sentence you can disprove Investor thesis fit is the reason a particular investor could rationally consider this particular company. It is not the fact that both websites mention AI, climate, or fintech. Write one sentence before researching names: “We sell [product] to [buyer] for [job], earn revenue through [model], and are raising [round] to reach [milestone].” If that sentence changes every time you find an attractive fund, you are editing the company to fit the list. A useful fit claim has two halves: a fact about the startup and a fact about the investor. “We automate hospital purchasing; the fund explicitly backs software sold to healthcare institutions” is testable. “They love innovation” is not. Build the startup half with the [investor search brief](/resources/build-an-investor-search-brief) before scoring the other half. ## Keep constraints separate from reasons to care Split your research into three questions. Can this investor consider the deal? Why might they want to? What remains unknown? Combining these into one flattering paragraph conceals decisive problems. A stated exclusion can outweigh several positive similarities. A medical-device fund may be enthusiastic about healthcare while excluding healthcare services. A first-check investor may be wrong for a business that has already completed several institutional rounds. Do not compensate for an explicit mismatch by adding more overlapping keywords. Equally, do not turn silence into exclusion. An investor that does not mention your exact workflow has not necessarily rejected it. Use “unresolved” when the evidence cannot support either conclusion. K9's published criteria illustrate why industry tags are insufficient: its requirements discuss revenue structure, founder location, and how early it invests. These are examples of fields to inspect, not a recommendation or a permanent description of its mandate. [K9 Ventures criteria](https://www.k9ventures.com/criteria/), checked September 24, 2026. Use the [AI seed investor list](/resources/investors/seed-ai), [health seed investor list](/resources/investors/seed-health) or [fintech seed investor list](/resources/investors/seed-fintech) to find stated mandates to test against that company sentence. Use the [seed-fund sector focus comparison](/resources/where-seed-funds-invest-sector-counts) to distinguish a broad category from a specific buyer and workflow. ## Read evidence at the right level Keep firm, fund, partner, and transaction evidence separate. A firm's growth vehicle does not automatically tell you what its seed vehicle can do. A partner's previous employer does not prove their present remit. An investment announcement proves something happened, not that every similar company is now eligible. Index's historical guidance recommends a targeted approach that considers the person as well as sector, stage, geography, and investment history. Use that as a research principle, not as a current statement of its terms. [Index Ventures, 2014](https://www.indexventures.com/perspectives/what-index-looks-for-in-an-investment/). Capture the source URL, publication date when available, observation date, exact entity, and the claim you think it supports. For each claim, ask what would make it false. A new role announcement might invalidate a partner map. A newer mandate might supersede an older thesis. A portfolio company's current product might change your conflict assessment. ## A fictional comparison Imagine HarborDesk, a fictional startup selling scheduling software to independent clinics. It charges a subscription and is raising a fictional $900,000 round to validate a repeatable sales motion. Fund A describes healthcare investing, but its supplied mandate is limited to clinical therapeutics. Fund B describes business software and has a recent, relevant clinic-operations investment. Fund C has no clear mandate but a partner previously worked in healthcare. The research outcomes differ: - Fund A: explicit mismatch under the supplied mandate. Do not prioritize unless newer evidence changes it. - Fund B: plausible thesis fit. Investigate the portfolio overlap and [check-size fit](/resources/check-size-reality) before outreach. - Fund C: an exploratory candidate. The biography is a reason to investigate, not proof of investment appetite. None is a funding prediction. Fund B can still decline. Fund C can still be interested. The exercise allocates research attention without pretending to know the investment decision. ## Define what changes the decision Use three working buckets: approach, clarify, and park. Approach means the important supplied constraints align and you can explain a specific reason to talk. Clarify means one answer could change the decision. Park means an explicit incompatibility or weak relevance makes other candidates a better use of time. For a clarify candidate, write one question, not five vague tasks. “Does this vehicle make new investments before commercial revenue?” is actionable. “Learn more about the fund” is not. Keep a small exploratory lane for candidates with incomplete evidence. Otherwise a richly documented but mediocre fit can crowd out a promising less-public investor. Explain the uncertainty instead of manufacturing a confidence percentage. ## The failure case: mistaking resemblance for permission A portfolio logo is tempting evidence because it feels concrete. Yet the investor may have entered at a later stage, through another vehicle, or before the company changed direction. It may also create a conflict. “They funded a similar company” belongs in the research notes, not automatically in the positive column. Before sending anything sensitive, use the [portfolio conflict review](/resources/portfolio-conflicts). Then map the appropriate human with [the partner map](/resources/the-partner-map). You can optionally start with a [Find investors for your startup](/try); a generated shortlist is a research starting point, not an endorsement, contact guarantee, or promise of funding.