Fundraising in the
age of AI.

Early-stage investing has changed…a LOT.

It’s not just you. Raising money right now is just…weird.

Fundraising advice is everywhere. Investors post deck tear-downs every week, accelerators publish templates, and an AI tool will produce a reasonable outline in under a minute. Most of it is correct on the fundamentals. Very little of it explains how much the seed market has changed in the last two years, and those changes affect how you should build your deck and run your raise.

I build investor decks with founders, and I also work with venture firms on the materials they use to raise their own funds from LPs. The LP work gives me a clear view of the pressure your investors are under right now, and that pressure shapes almost every question you’ll get in a seed meeting. The seed slide map on this page is the same document we start every early-stage deck engagement with. Below is the context I walk founders through before we fill it in.

Your investors are raising too

Every seed investor you pitch has investors of their own, and most of them are in a harder fundraising market than you are. According to the PitchBook-NVCA Venture Monitor, 73.1% of the capital committed to US venture funds in Q1 2026 went to five firms, and emerging managers continued to struggle to attract capital. Many of the smaller funds that write the bulk of seed checks fall in that second group.

Liquidity is still tight as well. PitchBook described the IPO window as mostly closed in the first quarter, so many funds raised in 2021 and 2022 have returned little cash to their LPs. A GP in that position has to show markups and a credible path to distributions in order to raise the next fund, and a large share of the remaining capital is reserved for follow-on and bridge rounds in the existing portfolio.

For a founder, the practical result is that each new seed check has to justify itself more clearly. Investors want to know whether your company can raise its next round on schedule and at a meaningful step-up, because their own next fund depends on companies that do.

AI changed the clock

The second shift is speed. In the LP materials I worked on this year, one of the charts that landed hardest compared how long software companies take to reach $100M in ARR. Data from Vantage.vc put the median AI company at roughly 12x faster than the SaaS median, with top-quartile AI companies getting there in about seven months. A GP I work with put it simply for his LPs: a milestone that used to take seven years can now take seven months.

[CHART: Time to $100M ARR, SaaS vs. AI. Source: Vantage.vc LP Intelligence Brief, April 2026]

Fast outliers reset expectations for everyone else. Seed investors now see companies with small teams shipping real product and signing customers on pre-seed money, and every deck gets measured against them. AI startups also captured roughly half of all pre-seed dollars in the first half of 2026, according to Carta, so competition for investor attention is concentrated in a few categories.

Two consequences come up in nearly every seed engagement I take on. The bar for what you built with your first check is higher, since AI tools let a small team ship far more than a small team could three years ago. And every founder, AI company or not, needs a clear answer on defensibility. For an AI company, the answer has to explain what you have that a model provider or a well-funded competitor can’t easily copy, such as proprietary data, distribution or deep workflow integration. For a company outside AI, the answer has to explain why AI makes your business stronger.

How to approach your seed round

With that context in mind, here is how I advise founders to approach the raise.

Know which kind of fund you’re pitching. A large multistage firm writing a seed check is often buying an early position in companies it hopes to lead later. A small emerging manager needs companies that can return a meaningful share of the fund and show a markup in time for its next raise. Angels and operator funds tend to focus on the founders and the problem. The same deck can work for all three, but each conversation should reflect what that investor needs.

Pitch the path to Series A. Your milestones slide should describe what this round will prove and when. Tie the round size to those milestones, show use of funds as percentages across product, go-to-market and hiring, and be specific about the evidence you expect to have in hand when you raise again.

Lead with evidence. At seed, traction can be revenue, pilots, design partners, signed letters of intent or a log of customer conversations. Count whatever you have honestly and show the trend. A short list of real customer conversations is more persuasive than a revenue projection with no customers behind it.

Show capital efficiency. Investors know how much a small team can build now. Show what you accomplished with the money you’ve already raised and how lean the next stage will be.

Build your market size from the bottom up. A top-down TAM pulled from an analyst report reads as desk research. Show the inputs: the number of target customers, your price, and the share you can realistically reach. AI tools will happily produce a TAM figure with no source behind it, and investors have seen plenty of them.

Write the map before you design the deck

Every founder wants to see slides. I understand the instinct, and I still hold off on design until the content is settled. A slide map is a plain document, with the slide name in one column and the headline, supporting points and visual notes in the other. Cofounders, early employees and advisors can all comment on the same page, and the discussion stays focused on the argument.

The mapping process also surfaces disagreements. At seed, the most common one is cofounders describing the company differently. One leads with the technology, the other leads with the customer, and the pitch changes depending on who is in the room. Investors notice when the story shifts between meetings, and settling it in a document is far easier than discovering the gap in a partner meeting.

A few rules we apply to every seed map:

• Put the vision before the product, and keep the product section short. At seed, an investor is underwriting the founders and the insight.
• Place the team slide right after the problem, so the investor knows who is making each claim. [LINK: founder-market fit post]
• Trace every number to a source, and leave unconfirmed figures as labeled placeholders.
• Pitch what exists today. Label demos, prototypes and roadmap items clearly, and leave out buzzwords your team doesn’t use when describing the product themselves.
• Make team credentials precise enough that a former colleague would agree with every word.

Build for the way decks get read now

Decks get read on phones, between meetings, often at thumbnail size, and many are summarized by an AI tool before a partner opens them. Put the point of each slide in the headline, keep all copy as live text so a summary tool can read it, and name your category explicitly in the first three slides.

At pre-seed, a strong team with a sharp insight can sometimes replace the deck with a one-pager covering the mission, the team, the insight, the bets you are making, progress, initial go-to-market, competition and market. The template includes that structure as an alternative.

Before you send the first deck

• Research each fund’s size, vintage and recent fundraising so you understand its pressures
• Define what this round will prove and what your Series A evidence will look like
• Fill in the slide map in a shared document before anyone opens a design file
• Build your market size from inputs you can defend
• Have each cofounder review the full map separately

Your investors are making a harder case to their own LPs than they were a few years ago. The seed decks that move forward are the ones that give them the evidence to make it.

Sources

Vantage.vc, LP Intelligence Brief, April 2026; Q1 2026 PitchBook-NVCA Venture Monitor; Carta, State of Pre-Seed: Q2 2026