The Founder-Market Fit Read: Earned, Plausible, or Borrowed

Before the first partner meeting you have to answer one question: does this founder have an unfair, earned advantage in the specific market they are building in, or just a plausible-sounding story? This guide is for early-stage investors, platform and talent partners, and angels who want to answer that from public evidence rather than from the pitch. It gives you named dimensions, a scoring method, the false positives that catch most people, and a three-band verdict - Earned, Plausible, or Borrowed - with the artifacts that back each one.

The existing founder reads grade commitment, buildability, and partnership durability. None of them grade earned advantage in the specific market, which investors say is the deciding factor at pre-seed and seed. That is the gap this fills.

What the Earned, Plausible, Borrowed read decides

This read decides one thing: whether a founder's claimed advantage in a specific market is earned, merely plausible, or borrowed from credentials that do not map to the actual job of winning that market. It is a judgement call you make repeatedly, before the first partner meeting, from artifacts you can pull without the founder's help.

The three bands are not a ranking of founder quality in general. They describe the relationship between the founder and this market:

• Earned. The advantage shows up in the public record. Tenure in the exact market, shipped work that solved the problem, writing that names a non-obvious constraint, a network of reachable buyers in the target segment. The founder could write the unfair-advantage sentence and the evidence agrees.
• Plausible. The story is coherent and the credentials are real, but the record is adjacent rather than exact, or the strongest evidence is narrative rather than shipped. You would advance to the meeting to probe, not to pass.
• Borrowed. The credibility is real but it belongs to something other than winning this market. A viral post, a prestigious employer, a conference presence, or domain depth aimed at the wrong buyer. From a distance it looks like fit. Up close the artifacts do not support it.

Why three bands and not a pass/fail? Because the useful output of a founder-market-fit read is rarely a rejection. It is a sharpened agenda. Plausible tells you exactly what to disconfirm in the room. Borrowed tells you the one thing that would have to be true for the story to hold.

The dimensions and the artifacts that prove them

Score five dimensions, and attach a specific public artifact to each one. Practitioner frameworks converge on roughly this set: domain expertise, execution ability, insight quality, network access, and authenticity of motivation. The discipline that separates a read from a vibe is the artifact: every score points at a thing with a URL, not at an adjective.

Here is what each dimension proves, the artifact that proves it, and what it looks like when it lies.

One source puts the specificity rule sharply: write "I ran ops at three companies in this vertical," not "I know the space." If a score cannot be stated that concretely, it is not a score yet.

A note on what the credentials actually predict. The data points away from the young-genius read. The mean founding age of the fastest-growing 1-in-1,000 US ventures is 45.0, and for patenting firms the mean is 44.6. Conditional on founding, a 50-year-old is 1.8 times more likely than a 30-year-old to reach upper-tail growth. The credential that matters is accumulated industry depth, not youth. Treat a seasoned, specific-industry founder as the statistical norm for upper-tail growth, and treat the reflex to discount age as a systematic false negative.

The origin test: problem-first or opportunity-first

Classify the founder's origin before you score, because it reframes every other artifact. Problem-first means the founder lived the problem and then went looking for the solution. Opportunity-first means they scanned the market, saw a big business, and picked it.

Brian Balfour frames the two as "I started experiencing X problem, so I looked for solutions and had the insight of Y" versus "I looked at the market and it's a huge business opportunity." Neither is disqualifying. But they weight the other dimensions differently. A problem-first founder's domain-expertise score should carry more, because the insight is likely earned. An opportunity-first founder has to earn the advantage some other way, usually through execution and network, and the bar for the unfair-advantage sentence rises.

You can infer the origin from the public record without asking. Build a dated career timeline and look for one of two shapes:

• Problem-first shape. The current market connects to prior employment or a lived pain that is visible in the record. The founder worked in the vertical, or built the thing adjacent to it, before starting up.
• Opportunity-first shape. Sequential, unrelated jumps across hot categories. Crypto, then NFTs, then AI, with no thread. The narrative is polished precisely because it has to carry weight the record does not.

There is no verified public number for the outcome gap between these two origins. Blogs circulate survival figures attributed to Stanford, CB Insights, and Startup Genome, but none of those is established in a primary source I can stand behind, so do not assert a number. Use the distinction qualitatively: it changes how you weight the rest, not what you claim about odds.

Scoring: the procedure

Here is the repeatable procedure. It runs about two hours of analyst time for the pull and scoring, plus partner time for the verdict and agenda. Do it before the meeting, in this order, though you may fold the origin classification into the scoring if you prefer.

The scale is deliberately coarse. Ulu Ventures scores first-meeting rubric elements as neutral, plus, or minus, and that granularity is honest about how much a public-artifact read can actually distinguish. A five-point scale invites false precision. Use plus, neutral, minus, and make the plus earn itself with an artifact.

Here is a scoring sheet you can copy.

Founder:
Market:
Origin (problem-first / opportunity-first): ___  Evidence: ___

Dimension        Score (+/0/-)   Artifact + URL
Domain expertise   ___           ___
Execution          ___           ___
Insight quality    ___           ___
Network access     ___           ___
Authenticity       ___           ___

False-positive checks (pass/downgrade):
 Buyer-identity match:   ___
 Trend-chasing pattern:  ___
 Names a hidden rule:    ___

Unfair-advantage sentence:
"This founder specifically wins this market because ___."

Verdict (Earned / Plausible / Borrowed): ___
Backing artifacts (2-3): ___

On weighting: there is no industry-standard founder-market-fit weight. Hustle Fund's published pre-seed scorecard puts insight at 10 percent and team at 25 percent; NFX refuses to number it at all and leads with four non-scored signals. The consensus is not a number. It is that the team-market pair dominates at pre-seed, because the company effectively is that pair when traction is still early.

The step that does the most work is the unfair-advantage sentence. It is a forcing function. If you cannot write why this founder specifically wins without reaching for generic filler, the plus scores are probably borrowed, and you have your answer before you tally anything.

Pulling the artifact set by hand takes 20 to 40 minutes per founder, and the pull is the step people skip when they are busy. Refolk returns the work-history, shipped-work, and network artifacts from public LinkedIn, GitHub, and the open web in one pass, so the evidence sheet is assembled before you start scoring rather than during.

How this read goes wrong

The failure modes all share one shape: a real credential standing in for an earned advantage it does not actually represent. This is the most valuable section, because the whole point of the read is to catch credential-without-insight before the meeting charms you out of it. Run every high score through these checks.

• Tenure mistaken for insight. Long industry tenure can mean absorbed conventional wisdom, not an earned secret. The ten-year veteran who recites incumbent assumptions reads as domain-expert and is not. Check: can they name a non-obvious market constraint that outsiders miss, such as a procurement cycle, a regulatory friction, or a switching cost? If the writing and the record contain only the obvious, downgrade.
• Consumer familiarity read as readiness. A heavy user of a product category is assumed to understand the business of it. Check whether the evidence is consumer familiarity or operator depth. Loving the product is not the same as knowing how it is sold, priced, and churned.
• Resume does not equal buyer. Strong credentials in the wrong persona. The clinician founder selling to hospital finance has real domain depth and the wrong buyer empathy, so the go-to-market will misfire. Check that the background matches the economic buyer, not just the domain.
• Trend-chasing disguised as vision. A polished narrative stretched across crypto, then NFTs, then AI. Check the dated career timeline for sequential unrelated jumps. The smoother the story over a discontinuous record, the more the narrative is doing load-bearing work the artifacts cannot.
• Repeat-founder halo. A prior exit raises odds to about 30 percent, not to certainty, and even repeat founders fail the majority of the time. Check that the prior exit is in the same market, not a generic serial badge. A logistics exit does not buy an earned advantage in fintech.
• Credibility theater. A viral post or conference presence substituting for shipped work. One source names it exactly: a slick resume, industry conference selfies, a Medium post that went mini-viral. Check for a shipped product, a repo, a patent, or a signed design partner, not just visibility.
• Network vanity. Follower counts standing in for people who would take the call. Check for named, reachable customers and partners in the target market.
• Score inflation from adjectives. "Knows the space" with no artifact. Check that every score cites a specific dated artifact, or revert it to neutral.

That geography point has a second edge. Refolk's index shows UK fintech founders clustering at a few shared employers while US fintech founders are dispersed. Shared-alumni markets inflate perceived insider status, which is exactly the condition that produces Borrowed verdicts dressed as Earned. In a market where everyone trained at the same handful of incumbents, pedigree is the ambient default, so the earned secret has to be visible above it.

Mapping scores to the verdict

Map the surviving scores to one of three bands using the shape of the evidence, not a point total. A verdict is defensible when you can name the two or three artifacts that put it where it is.

Read the matrix this way. Exact-market plus shipped evidence is Earned. Exact-market but mostly narrative, or adjacent-market but shipped, lands in Plausible, and the band tells you which gap to probe. Mostly narrative in an adjacent market is Borrowed, and the useful output is the single artifact that would move it: a signed design partner, a repo, a named reachable buyer.

Anchor the bands against the base rate. Previously successful founders succeed about 30 percent of the time, previously failed about 22 percent, and first-timers about 21 percent.

The spread between the best and the worst case here is nine points. That is the ceiling on how much any read can move the odds, and it is why an Earned verdict should raise conviction and sharpen the agenda rather than close diligence. The most dangerous move is to let Earned end the inquiry.

An Earned verdict buys you nine points of base rate, not a decision. It sharpens the agenda; it does not close it.

The founder population also tells you how much of this work is a sourcing problem versus a judgement problem. In Refolk's index there are 17,650 current fintech founders in the US and 13,430 in US hospitals and health care, which means in a deep market the constraint is the read, not the supply.

Before you call it done

Run this checklist before you send the verdict into the partner meeting. Each item is a thing to verify, not a topic to consider.

To keep the read current, re-pull the artifact set rather than trusting a sheet from a prior cycle. Work history, shipped work, and writing all move, and a founder who was Plausible six months ago may have shipped the thing that makes them Earned, or jumped categories in a way that makes them Borrowed. The verdict is only as good as the evidence sheet behind it, and the 5.9x-versus-1.1x diligence gap says plainly that the hours you spend assembling that sheet are the hours that drive the outcome. The read is not the opinion. It is the artifacts, and you can always check them again.