The Founder-Market Fit Score: Grading a Founder's Edge From Public History

Deciding whether a founder actually fits the market they are building in is one of the few judgement calls a seed investor makes over and over, and usually makes on instinct. This guide is for early-stage investors, platform and talent partners, and angels who want to convert a gut "I like this founder" into four graded, sourced sub-scores before the founder is ever in the room to spin the story. It gives you a scoring instrument: the dimensions that matter, the public source that proves or fakes each one, and the arithmetic that turns four sub-scores into an advance, dig-deeper, or pass verdict two people would reach independently.

Most of what ranks online for founder-market fit is founder-facing: advice on how to perform fit in a pitch. This is the other side of the table. It is built to be run against a real founder from public history alone, before diligence calls, so your read is anchored in what you can reconstruct rather than what you are told.

Why the founder score moves the valuation more than the idea

The founder is the heaviest single lever in the most-used angel instrument, so a defensible founder-market fit read changes the number, not just the conviction. The Bill Payne Scorecard, formalized by angel investor Bill Payne, weights entrepreneur, team, and board at 30%, more than the size of the opportunity (25%) and more than product or technology (15%).

The mechanism is simple. At pre-revenue there is almost no other evidence to price. Revenue is thin or absent, the product may be a prototype, and the market is a projection. So the founder factor absorbs disproportionate weight, and a soft "I liked them" is doing real work in the valuation whether you admit it or not. That is the argument for a scoring instrument: it makes the heaviest input auditable.

The frameworks in circulation do not agree on how many dimensions to score, but they converge on which ones matter. Four recur across independent investor sources.

How the named frameworks line up

No two published frameworks name the same dimensions, but they overlap heavily once you strip the labels. This is the field, side by side.

I score four rather than six because six splits hairs that public history cannot reliably separate: lived experience and domain expertise collapse into one signal when all you have is a timeline, and personality is not reconstructable from public records without guesswork. Four dimensions, each tied to a distinct public source, is the most you can grade defensibly before the meeting.

What each dimension proves, and what it looks like when it lies

Every dimension has a public source that establishes it and a way that source misleads you. The discipline of the score is holding both in view: what the signal proves, and what it looks like when it fakes.

Domain expertise is established by employment history, education, and projects. The scorecard practice is to list the founders' profiles up to incorporation, noting qualifications, years of experience, and key relevant companies. Employment history proves tenure and names real employers. What it does not prove is depth. Presence is not proximity. Kela's example is the sharpest statement of this: a founder who built high-frequency-trading systems at a quant fund and now builds compliance tools for trading desks has fit, while a founder who read about trading and built the same tool does not, and customers tell the difference in the first ten minutes. Proximity beats prestige because lived experience encodes buyer nuance a resume cannot.

Authentic obsession is established by unpaid, pre-company activity, not by anything said in the room. NFX's litmus test is whether the founder would choose to work on the idea in their free time, work effortlessly on product and customer issues, not notice time passing. In public history that shows up as side projects, open-source contributions, writing, and community participation predating any raise. It lies as obsession theater: trend-following dressed as mission, which reads as authentic passion in person and evaporates under a timeline. Behaviour before any financial upside is costly to fake, which is exactly why unpaid history outranks pitch enthusiasm.

Proprietary insight is established by a non-obvious claim that traces to lived experience rather than market research. Seafund's framing is that the most reliable historical signal is where the founder spent time and what they learned from failures in the domain, alongside a proprietary insight, a non-obvious thing they know that most do not, which usually traces back to lived experience. It lies as consensus: a confident thesis anyone in the sector already holds, delivered with enough conviction to sound proprietary.

Network in the market is established by named, reachable first-degree contacts in the target ecosystem. It lies as connection count: a large LinkedIn network read as go-to-market capacity when none of it is actually the buyer.

The weighting instruments, and how much to trust them

Two named instruments assign explicit weights, and they disagree on how much of the total the founder should carry. Use them as reference points, not gospel; each was built for a different job.

The Payne Scorecard is a valuation tool. It multiplies your region's median pre-money valuation for similar-stage startups by a weighted seven-factor scorecard, scoring against a regional baseline where 1.0 equals the median. Its 30% founder weight is one line in a broader valuation. The FormLine quiz is a fit tool, and when you sum its domain expertise (25%) and lived experience (25%) it puts roughly 50% on founder-related factors, with thresholds of hot at 80, warm at 60, and cold at 40. The Berkus method takes a different tack again, assigning up to $500K per success factor across five categories including team, for a maximum pre-money of about $2.5M.

My recommendation: score the four fit dimensions on their own 1-5 scale, then, if you use Payne, let the fit read drive whether you push the 30% founder factor above or below the 1.0 baseline. Keep the two exercises separate so the valuation math does not contaminate the fit read.

Reconstructing the timeline is the slow part of this job, and the part where a good index earns its place. In Refolk's index, industry-code-filtered founder queries returned zero matches while a "fintech" headline keyword surfaced over 2,000 US founders, because founders self-describe by problem space in their headlines rather than by industry code. Searching Refolk by the language founders actually use is the higher-recall path to the raw history the score depends on.

The scoring procedure, step by step

Run these seven steps in order. There is a documented order dispute worth naming: NFX leads with obsession, while scorecard practitioners lead with employment history. I lead with the raw pull so that whichever dimension you weight most, you are scoring it against the same complete timeline.

The whole pass runs in about two hours: 30 to 60 minutes on the timeline, then 15 to 20 minutes on each dimension, 15 on the red-flag pass, and 10 to blend. If two reviewers run it on the same founder and land more than one point apart on any sub-score, the disagreement is almost always a citation gap in step one, not a judgement gap.

Turning four sub-scores into a verdict

Score each dimension 1 to 5. The verdict is not a simple sum, because a fatal weakness on one dimension cannot be bought back by strength on another.

Domain expertise:   __ / 5   (role touched the buyer? Y/N)
Network in market:  __ / 5   (named first-degree buyers/hires: ______)
Authentic obsession:__ / 5   (unpaid pre-company activity dated: ______)
Proprietary insight:__ / 5   (non-obvious claim, sourced to: ______)

Verdict:
- ADVANCE      = domain >= 4 AND obsession >= 3 AND no unresolved red flag
- DIG DEEPER   = total >= 12 but one dimension unproven from public history
- PASS         = domain <= 2 with no buyer proximity, OR any live red flag
Written rationale (2-3 sentences, cite the timeline): ______

How this read goes wrong: the failure modes

This is the most valuable part of the instrument, because every dimension has a false positive that feels like a strong signal. Log each one against the founder in front of you.

The one the frameworks underprice is over-expertise. Deep incumbent experience gets scored as a pure asset, but the same investors who reward domain knowledge call it a double-edged sword that can hinder out-of-the-box thinking and become baggage. Peer-reviewed work confirms it: accumulated experience rigidity obstructs the unlearning of outdated methods, what Fisher and Keil term the curse of knowledge, and a 2013 study found over 80% of expert radiologists missed an inserted gorilla image, an illustration of expert tunnel vision. Score for adaptable expertise, not routine expertise. The tell is "this is how it has always been done" framing in the founder's public writing.

Two structural checks close the pass. Co-founder conflict is described as the number one killer of early-stage startups, so detectable co-founder history matters even when the lead founder scores well. And treat the youth-bias instinct with suspicion: Wadhwa's studies of 549 entrepreneurs across 12 high-growth industries and 502 companies found the average and median age of successful founders was 39, with twice as many over 50 as under 25. "Too experienced" is rarely the real failure mode.

Where the founders are, and how to read the geography

Founder-referencing signal is keyword-borne, not industry-code-borne, which changes how you reconstruct history at scale. In Refolk's index, industry-code-filtered founder queries returned zero matches, while a single "fintech" headline keyword query surfaced at least 2,000 US-based profiles titled Founder or Co-Founder, with the count capped at 2,000.

A small geographic sample from that population points to hub concentration rather than precise shares. Treat it as directional.

This sample rests on 25 profiles drawn from a 2,000-capped total, so reading precise market shares into it would be a false positive of its own. What it supports is a claim about where to look first, not a market map. The practical takeaway: search by the problem language founders put in their headlines, and expect fintech founder density to cluster around NYC, Chicago, and the Bay Area.

Before you call the score final

Run this checklist before you write the verdict. Each item is a thing that, if missing, means the score is not yet defensible.

Keeping the instrument honest over time

The score is only as good as the history behind it, so the maintenance job is keeping the reconstruction path current. Two things drift. First, how founders self-describe: the reason to search by headline keyword rather than industry code is that founders label themselves by problem space, and the vocabulary of a space moves. Re-check your query language against how founders in the sector actually write their headlines, not against a taxonomy. Second, the unverified stats that circulate in this field: before any of them enters a verdict, trace it to its primary source, because a score built on borrowed numbers inherits their errors.

Run the four-dimension pass the same way every time and it does the thing a framework is supposed to do: two partners score the same founder from public history and land in the same place, before anyone has heard the founder tell the story. That is the difference between a scoring instrument and a first impression.