What Investors Look for in a Startup at Seed Stage (and What They Are Trying to Prove Is Real)

Seed investors look for five signals: team, market timing, traction quality, unit economics, and narrative. Across 885 institutional VCs surveyed by Gompers, Gornall, Kaplan and Strebulaev in 2016, the founding team was named an important factor by 95% of firms, ahead of business model (83%), product (74%), market (68%) and industry (31%). The five signals below are how those categories get tested in a seed meeting, read from seven years on the investor side: business model splits into traction quality and unit economics, and narrative is the consistency test applied across all of them. Every question at seed tests whether one of those signals is real.

Last updated: September 15, 2026. Every figure re-checked at its primary source, Carta's mid-2026 seed and Series A medians added, and a verdict on which founders this guide serves.

This guide breaks down each signal, what counts as proof at seed, how the bar shifts at Series A, and the practical checklist founders run before opening outreach. Written by David Rakusan, a founder with seven years on the investor side.

What Seed Investors Are Actually Trying to Determine

The first investor meeting is a proof mechanism that looks like a conversation.

A seed investor is trying to determine, in 30 to 60 minutes, whether the founder's claims about themselves, their market, and their early data are real. The slides are a starting point. The conversation is where conviction actually forms. According to the survey by Gompers, Gornall, Kaplan and Strebulaev, covering 885 institutional venture capitalists at 681 firms, the management team was named an important factor in pursuing deals by 95% of VC firms and the single most important factor by 47%. Business model was named by 83%, product by 74%, the market by 68% and the industry by 31%. Team outranks every other category, so the founder is read first and everything else is weighed against that read.

What the founder is being tested on is judgment. Can this person see the market clearly. Can they describe the customer accurately. Can they tell the truth about what is not working yet. Can they hold up under stress without going defensive. Those tests run underneath every question, and the answer is built from cumulative evidence across the conversation rather than from any single slide.

This is the reason the same founder gets different reads from different investors in the same week. The startup did not change. The lens did. A 2023 study in the Journal of Small Business Management, built on case studies of 15 VCs, identified five distinct approaches VCs use to size up teams, from a purely intuitive approach to a scientific rational one, and describes the mix as heterogeneous: different combinations of subjective judgment and objective data. Different investors bring different proxies for what a great founder looks like, and the founder cannot control which proxy is being used in any given meeting.

Who This Guide Is For

This guide is for a first-time founder who wants to know what investors will push on before the first meeting, and who is tired of repeating the same twelve answers in every call.

If you are a repeat founder with a warm network and a prior exit, the funnel below still applies, but the signals get read faster and this guide works as a checklist. If this is your first raise, read it as the map of the first three meetings. If you are not raising yet, run it as a twelve-months-out diagnostic. Finding a weak signal now is cheap. Finding it in meeting one costs you the meeting.

Why Paper Fails to Prove These Signals

A pitch deck is a curated story. Financial models are assumptions packaged as numbers. Reference lists are hand-picked. Investors know this. Their job is to find what the deck cannot show.

The data on how investors actually consume decks makes this concrete. DocSend's 2024 seed and pre-seed reports, built on more than 400 fundraising startups, found investors spent 40% more time on seed-stage Team slides and 30% more on pre-seed Team slides in 2024 than the year before, while market context and competition lost attention. The signal investors are most hungry for is the one a static deck shows least well.

Paper fails because it is static. It cannot answer a follow-up. It cannot show how a founder thinks under pressure. It cannot link a revenue claim to a live Stripe dashboard. It cannot link a hiring story to actual LinkedIn histories. It is a snapshot, and snapshots can be staged. Conviction forms in live conversation because that is where the founder is forced to defend the snapshot in real time.

There is a structural reason the proof gap exists. A 2023 paper in the Journal of Business Venturing Insights on the inefficiencies of venture capital funding frames the core friction as information asymmetry between founder and investor, and notes that both sides deal with it with very different tools. Every new fund reconstructs its own map of that asymmetry from scratch. Conviction cannot be transferred from one fund to another. That is why a founder pitches the same numbers 20 times in a row and still gets asked the same opening questions in meeting 21.

The Five Signals Investors Weight at Seed

Here are the five signals in the order most seed investors weight them, with what each one actually tests and what counts as proof.

1. Team

What investors are testing: founder-market fit, judgment under uncertainty, and ability to recruit. Pedigree matters for an indirect reason. Investors do not weight a Stanford degree because the degree itself is intrinsically valuable. They weight it because it is a predictable screen the founder already passed, which reduces one variable they have to size up themselves.

What counts as proof at seed: a clear answer to "why are you the right person to build this, in this market, right now." Specificity wins. A founder who can describe the exact pain they lived through, in the exact words their target customer would use, is more credible than one who describes the market in TAM language. Dan Gray, writing for Crunchbase News, notes that both the best- and worst-performing founders tend to be first-timers, citing research by Gompers, Kovner, Lerner and Scharfstein; 49% of unicorn founders had started a company before, which leaves a majority who had not.

2. Market

What investors are testing: timing and size. Timing is more important than size at seed. A $10B market with the wrong timing is dead capital. A $1B market with the right tailwind compounds.

What counts as proof: a clear thesis on why now, supported by either a technology shift, a regulatory shift, or a behavior shift. Paul Graham's essay How to Get Startup Ideas compresses the timing test into one line: "Live in the future, then build what's missing." If you cannot explain what shifted in the last 18 months that makes your startup viable today, your market thesis is weak.

3. Traction

What investors are testing: whether the early data is real, recurring, and unsubsidized. A spike from a launch is not traction. A 12-week retention curve is.

What counts as proof at seed: in my seven years on the investor side, the bar was usually 6 to 12 months of one core metric trending in the right direction, and the bar keeps rising. Waveup's 2025 fundraising study, a survey of 56 VCs, found round expectations jumped one full stage: pre-seed now expects seed traction, and seed plays by Series A rules. The same survey named net revenue retention as the new big one among the metrics VCs want to see. The gap to the next round is longer too. Forum Ventures' study of 300+ pre-seed and seed B2B SaaS deals found the average time between seed and Series A stretched to more than two years in 2024, up from 1.7 years in 2019, so the seed metric has to keep compounding through a long gap. The cleanest founder-side proxy for product-market fit is the one Sean Ellis built from benchmarking nearly a hundred startups: ask users how they would feel if they could no longer use the product, and measure the share who answer "very disappointed". Writing in First Round Review, Rahul Vohra reports that Ellis found the magic number was 40%, with companies that struggled to grow almost always landing below it and companies with strong traction almost always above. Which metrics carry that weight is covered in the seed metrics investors actually track, and founders without revenue yet can still build the case, as our guide to proving traction pre-revenue shows. The cost of getting this signal wrong is documented: CB Insights' 2026 review of 431 VC-backed failures found poor product-market fit cited in 43% of post-mortems, the most common root cause behind the money running out.

4. Unit Economics

What investors are testing: whether the business model holds up at scale. Gross margin, payback period, and retention. At seed, the standard is lower because the data is thin, but the question is the same.

What counts as proof: an honest model that shows how the unit economics evolve as the business grows, with clear assumptions and a stress test on the failure cases. The number seed investors increasingly anchor on is the burn multiple, net burn divided by net new ARR. David Sacks at Craft Ventures, who popularized it, calls a 2x multiple reasonable for an early-stage startup and 5x terrible. Below meaningful net new ARR the ratio is not computable, so at seed the useful version is the direction of travel and the assumptions behind it. Scale Venture Partners' benchmark of several hundred private and public SaaS companies puts the average at 3.4x for companies under $1M of ARR and about 1.6x across the whole dataset, and finds that the fastest-growing companies in each band are also the most efficient. The same CB Insights analysis found unsustainable unit economics behind 19% of failures. A seed founder who can explain the path from a 3.4x early multiple toward the 1.6x range gives investors a defensible answer for the Series A conversation.

5. Narrative

What investors are testing: consistency. The story you tell in meeting one needs to match the story your data tells, the story your team tells, and the story your customers tell. A great narrative answers the next investor's question before they ask it. A weak narrative invites partners to find inconsistencies.

What counts as proof: a 60-second version of the business that survives three rephrasings. The investor will pose the same question three different ways across one meeting to see if the answer holds.

How the Bar Shifts From Seed to Series A

The signals do not change between seed and Series A. The weighting does. This table shows how the same five signals get measured at each stage.

Peter Walker at Carta published benchmarks in July 2026 from more than 1,000 software rounds raised in the preceding six months: the median seed round raised $4.1M at a $24.3M valuation for 18% dilution, and the median Series A raised $14.4M at $80M, also for 18%. His own caveat travels with the numbers: medians are the middle of a wide range, and startups raising in the top decile at seed do not always raise in the top decile at A. Founders who walk in with proof of all five signals price at the top of the band. Founders who walk in with two or three of the five price at the floor, or do not get priced at all.

The Repetition Trap That Wastes Everyone's Time

Every fund starts from zero. Even with a warm introduction, the investor on the other end still has to form their own conviction. That conviction cannot be transferred from one partner to another, or from one fund to another. The result is the same questions, in the same order, in every meeting.

The aggregate cost is enormous. The same Gompers survey found VCs spend an average of 22 hours a week networking and sourcing deals, and that for every deal a firm closes it considers about 101 opportunities, meets management at 28, and takes 10 to a partner meeting. Those numbers describe what actually sorts deals. The five signals are the language investors use to justify a decision that sourcing, pacing and social proof have already shaped, which is why a founder can pass every signal test and still never reach the room. Proof and access are separate problems. Affinity's survey of nearly 300 private capital dealmakers found 85% now use AI to automate daily tasks, up from 76% the prior year. The meeting itself is still the proof mechanism. The AI speedup compresses the screening step. Conviction still forms in the room. On the founder side, the Waveup survey noted earlier found raise cycles stretch over 12 to 18 months for most founders, with 3 to 6 weeks for top companies. In my experience the same answers get repeated across 15 to 20 investor meetings because proof has not been established in advance. The repetition is structural.

For founders, the practical effect is that meeting one is rarely the one that closes. The first meeting usually ends with a request for more information: the deck, the financial model, the customer references, a follow-up call to dig into the founding story. By meeting two or three, the partner is rebuilding the same proof a different partner at a different fund already built last week. The founder is repeating themselves. The information loss between touchpoints is total.

The Proof Layer That Breaks the Loop

The fastest seed rounds happen when investors arrive at meeting one with the basics of all five signals already answered. Team background, market thesis, traction snapshot, unit economics, narrative consistency, all visible in a structured form the investor can read before the call. That is the proof layer.

For a founder on SeedForge that layer is the living profile described above. The 30-minute session covers the five signals the way a seed partner covers them, follow-ups included, and the answers, the deck and the documents sit at one link. The founder decides who sees it and can update it as the numbers move, so an investor who looks again a month later sees the business as it is today rather than the deck from the last raise. Completing the profile also unlocks a matched investor list with a drafted introduction per partner; the founder approves every message before it goes out, whether they send it themselves or let SeedForge run the outreach from their own LinkedIn. The profile and the first 30-minute session are free at seedforge.com.

This is the specific gap for "what investors look for." The investor wants proof that the five signals are real, before meeting one and in a form they can read in ten minutes. The proof layer handles the structured information gathering. Investors bring the judgment.

Three Screens Seed Investors Run That Never Reach the Deck

There are three screens most seed investors run that rarely make it into the deck conversation. They show up in passing questions, side comments, and post-meeting partner debriefs.

The first is reference quality. A reference call that contradicts a small claim from the founder ("we worked together at X for two years" turns into "he was at X for six months and we overlapped briefly") is fatal at seed. The investor will not say so. They will simply not move forward. The fix is to call your own references before pitching and align the story explicitly.

The second is decision velocity. Seed investors are increasingly using how quickly a founder makes small decisions during the conversation as a proxy for how they will make hard decisions later. A founder who hedges on every question signals fragility. A founder who picks a direction and defends it, then revises when given new information, signals judgment. The signal is read in seconds.

The third is incentive alignment with the cap table. A seed investor reading a cap table with three large angel cheques at very different prices is reading a future Series A problem. Carta's bridge-round data for Q2 2025 shows 16.6% of all cash raised on its platform came through bridge rounds, up from 11.8% a year earlier, with the highest share at Series A at 22.5%, which means many seed startups are already on a second tranche before their first priced round. Investors notice. Founders who can explain their cap table cleanly in one minute walk into priced rounds faster than founders who hand the investor a tangle.

Practical Checklist: What to Prove Before You Open Outreach

Before you start pitching, run this checklist against your own startup. If any answer is weak, the meeting will expose it. Fix it before the meeting, not during it.

Team: - Can you state your founder-market fit in one sentence, in your customer's words? - Do your two best references match the story you tell about yourself? - Are key hires either in place or named by role with a credible plan to recruit?

Market: - What shifted in the last 18 months that makes this startup viable now? - What is the credible expansion path beyond your wedge, the first narrow use case you win?

Traction: - Do you have 6+ months of one stable metric trending in the right direction? - Is the metric measurable, recurring, and unsubsidized? - Can an investor check it against an outside source (Stripe, GitHub, analytics, public reviews)?

Unit Economics: - Do you have an honest model with defensible assumptions? - Have you stress-tested the failure cases out loud? - Can you name your current burn multiple and the two levers that move you from roughly 3.4x toward 1.6x?

Narrative: - Record your 60-second version three times across one day without replaying the earlier takes, then compare the transcripts. What changed is what is not yet settled. - Does the data tell the same story your deck does?

For more on how investors press on these signals in live meetings, see our breakdown of what VCs actually ask in the first 3 meetings and how to prepare for a VC meeting once those signals are in place. If you are already mid-fundraise, our guide to seed-stage due diligence covers what investors look at after the first meeting.

Every question above is one the 30-minute session on seedforge.com asks in the order a seed partner asks it, and the weak answers come back named.

How This Question Is Answered Across Sources

Founders searching this question land on different doors depending on the phrasing. The table below shows how a few authoritative sources frame the same five-signal answer, with the angle each one leads with.

Each piece is one face of the same proof problem. The five signals stay the same. The format the founder uses to demonstrate them depends on the moment.

Frequently Asked Questions

What do investors look for in a startup at seed stage?

Seed investors look for five signals: team, market, traction, unit economics, and a narrative. In the Gompers survey of 885 VCs, the founding team was named an important factor by 95% of firms and the most important by 47%, ahead of business model (83%), product (74%), market (68%) and industry (31%). The other signals are weighted against the team signal.

What signals matter most at seed compared to Series A?

At seed, team carries more weight because traction is thin and the bet is on judgment. By Series A, unit economics and growth rate move to the top because the data exists to assess them. Carta's July 2026 benchmark, covering the preceding six months, puts the median seed at $24.3M on $4.1M raised, and Series A at $80M on $14.4M raised.

What counts as real traction at seed stage?

Real seed traction has three properties: measurable, recurring, unsubsidized. A spike from a Product Hunt launch is not traction. Monthly active users with a stable retention curve is traction. Waveup's 2025 survey of 56 VCs found round expectations jumped one full stage: pre-seed now expects seed traction and seed plays by Series A rules.

How important is founder-market fit at seed stage?

Founder-market fit sits inside the team signal, the one seed investors weight most heavily: in the Gompers survey of 885 VCs the founding team was named an important factor by 95% of firms and the most important by 47%. It answers one question: why are you the right person to build this, right now?

Do investors care about a polished pitch deck at seed?

A polished deck cannot rescue weak signals. DocSend's 2024 seed and pre-seed reports, built on more than 400 fundraising startups, found investors spent 40% more time on seed-stage Team slides and 30% more on pre-seed Team slides than the year before, while market context lost attention. Investors look for proof of the founder behind the deck.

What kills a seed-stage investment before the second meeting?

Three things end most second meetings early: a team signal that fails reference checks, a traction story subsidized by paid acquisition or one large customer, and a narrative inconsistency between what the founder said in meeting one and what the data shows. Each is a proof failure, and each surfaces in the follow-up the investor runs after the first call.