How seed VCs actually decide

LAST REVIEWED 2026-08 · SOURCED FROM 3 SESSIONS, APR–MAY 2026

TL;DR

Three GPs from three different fund types told the same story with different weights: pre-product funds index almost entirely on the founding team; deep tech funds look for a moat you can actually own (data, not algorithms or compute); traction-oriented seed funds want revenue calibrated to your ACV, plus renewal signal. Talk to 20–30 investors, not 3. Get introduced through their portfolio founders. And diligence them as hard as they diligence you.

The three lenses

The pre-product micro-fund (first checks $25–100K, decisions in 1–2 weeks): the bet is the team, almost entirely. What they fish for is resilience — backgrounds that prove someone doesn’t quit: marathons, collegiate athletics, years of unglamorous grind. The phrase used in the room: founders who “won’t give up even if they end up homeless.”

The deep tech seed fund (~$1.5M first checks): the bet is defensibility. Of the three moat ingredients — algorithms, compute, data — data is the only one a startup can realistically own. The mental model for stages: seed validates product, Series A validates market, Series B validates growth. The first milestone isn’t broad appeal; it’s a small set of users who genuinely cannot live without you.

“It’s much better than… trap them, then please them. Right? Trap them in the correct way.”

— Managing partner, deep tech seed fund (~$300M AUM) · session, Apr 2026, on building indispensable products

The traction fund (seed to Series A, outside the Bay): the bet is evidence. Their bar as stated: roughly $400–500K revenue, calibrated to ACV — fewer expectations for a $10K-ACV product, more for $50K B2B — with about a year in market and early renewal/NRR signal. Churn isn’t held against you at year one; CAC/LTV is a Series A conversation.

The honest takeaway for founders: know which lens the fund you’re pitching uses, because the same deck can’t serve all three.

Running the top of your funnel

Volume matters. The consistent advice: founders talk to far too few investors. Twenty to thirty targeted conversations, not three.

The warm-intro cheat code: VCs almost never turn down an introduction from one of their own portfolio founders. Find a portfolio company you can reach, get to know the founder, and come in through them.

Never open with the pitch. Ask qualifying questions first: how they got into venture, what their thesis is, and — the one almost nobody asks — where they are in fund deployment. A fund at the end of its deployment period isn’t writing your check no matter how good the meeting feels. End non-fit meetings early and ask for referrals; you earn future meetings by respecting time.

Read the polite no. “That’s great, keep me in the loop” is a no. Treat it as one and move on — then prove them wrong with material-change updates (see Running the raise).

Diligence goes both ways

“You should be doing as much diligence on your investors as they’re doing on you… if they’re gonna be a headache on the first call, it’s only gonna get worse.”

— GP, pre-seed micro-fund · session, Apr 2026

A troublesome investor sits on your cap table for roughly a decade. Ask for references from their portfolio founders — one GP noted only 3 or 4 of ~75 funded founders had ever asked him, and the ask itself signals sophistication. Scope it sensibly: reference-check leads and large checks, not every $10K angel.

What kills interest

• The single-big-logo strategy. A $100K pilot from a $100B company is play money for them — slow, secretive, and killed by any strategy shift. Mid-size customers give faster, more honest feedback and read as stronger evidence.
• Unprepared depth. The deep tech GP looks for founders who are two or three steps ahead of every question. Cheap fix from the room: run your own deck through an LLM as a skeptical investor before any human sees it — it catches most of the easy problems.
• Uncalibrated asks. Standard pre-seed is now around $1M. “AI” justifies speed, not an extra zero.

Numbers from the room

Figure	Value
Micro-fund first checks	$25–100K, decision in 1–2 weeks
Deep tech seed first check	~$1.5M
Traction fund seed bar	~$400–500K revenue, ~1 year in market, calibrated to ACV
Investor conversations to plan for	20–30+
Standard pre-seed round	~$1M
How long an investor is on your cap table	~10 years

Sources

Three sessions with fund GPs: a pre-seed micro-fund GP (Apr 2026), the managing partner of a deep tech seed fund (Apr 2026), and a partner at a mission-driven seed fund (May 2026).