Skip to main content
Data Room

The Pre-Seed Data Room: Five Documents That Unblock a First Check

7 min read
Dariia Vasylieva
Dariia Vasylieva
Founder, RaiseTalks

At pre-seed, a data room needs five documents: the deck, a simple financial model, a clean cap table, your founder agreements, and proof that your IP is actually assigned to the company. That is it. Everything else is optional and most of it is a distraction.

Founders consistently build too much at this stage, which is expensive in two ways: it takes weeks you do not have, and a bloated room buries the five things an investor actually needs.

Key takeaways​

  • Five documents. Deck, model, cap table, founder agreements, IP assignment.
  • In our experience the most common fatal gap is IP assignment, and it is the most expensive to fix late because it requires signatures from people who have already left.
  • A clean cap table that reconciles to every SAFE and note is worth more than a beautiful model.
  • Your model does not need to be right. It needs to show that you know which three assumptions drive the business.
  • Add documents only when an investor asks twice for the same thing.

Why five and not fifty​

At pre-seed, there is very little to diligence. Usually no revenue, sometimes no product, rarely a contract worth reading. The investor is underwriting a team and a thesis.

So the job of a pre-seed data room is not verification. It is removing reasons to hesitate. Every unanswered question is a small reason to wait and see. Five documents, complete and current, remove almost all of them.

If you want the broader field-level structure that this grows into, see the startup data room, field by field. At pre-seed, start here.

1. The deck​

Ten to fifteen slides. Problem, solution, product, market, business model, traction or evidence, team, competition, the ask, use of funds.

What makes a pre-seed deck work is not design. It is that every claim has a visible basis. A market size with a derivation. A traction number with a definition. A competitive claim naming actual competitors. Investors at this stage are not checking whether you are right. They are checking whether you know the difference between a claim and an assumption.

Export it as a PDF. Not a link that expires, not a Figma file, not a Canva share that requires a login.

2. A simple financial model​

Three years, monthly for the first eighteen months. Revenue build, headcount, costs, cash, runway.

Nobody believes a pre-seed model's year-three revenue. They are reading it for three things:

  • Do you know your drivers? The three or four assumptions that actually determine the outcome.
  • Is the burn plan coherent? Does headcount match the plan, does the plan match the raise?
  • When do you run out? Stated plainly, with the date visible.

A clean two-tab model beats an elaborate one. Elaborate models at pre-seed signal time spent in a spreadsheet rather than with customers.

3. A cap table that reconciles​

This is where most pre-seed rounds hit friction, and it is entirely avoidable.

Your cap table must show: every shareholder with share counts and percentages, the option pool (allocated and unallocated), and every SAFE and convertible note with its cap, discount and amount.

The test is simple: can an investor take your cap table, add the round they are considering, and get an ownership number that matches yours? If there is any ambiguity - an unrecorded SAFE, a verbal promise to an advisor, a pool that does not add up - they will find it, and it converts a funding conversation into a cleanup conversation.

Advisor equity handshakes that were never papered are a frequent source of this problem.

4. Founder agreements​

Founders' agreement or equivalent, with vesting.

An investor looking at a two-founder company with no vesting sees a company that loses half its equity to a person who left, with no mechanism to recover it. That is not a paperwork gap, it is a live risk to the asset they are about to buy into.

Standard shape: four-year vest, one-year cliff, with the clock ideally starting at the point real work began rather than at incorporation.

5. IP assignment - the one that actually kills rounds​

In our experience this is one of the most common gaps in early-stage diligence, and the most expensive to fix late.

Every person who wrote code, designed the product or contributed materially to the IP must have assigned that work to the company in writing. Founders. Employees. Contractors. The friend who built the first prototype over a weekend two years ago and has since moved to another country.

Why it is fatal rather than annoying: if your IP is not assigned, the company does not own the thing the investor is buying. That cannot be papered over and it cannot be fixed by the company alone - it needs a signature from a specific individual who has no obligation to give you one, and whose leverage increases the moment they learn a round depends on it.

Fix it before you raise. It costs a template and a few emails now; it costs weeks and sometimes a round later.

What to add only when asked​

Hold these back until an investor asks, and add them permanently once two investors have:

  • Incorporation documents and good standing certificate
  • Customer contracts or letters of intent
  • Detailed product roadmap
  • Technical architecture documentation
  • Employment agreements
  • Bank statements and accounts
  • Trademark filings
  • Prior round documents

The pattern to watch: if the same question arrives twice, it is not a one-off. It belongs in the room.

How to organise it​

Five documents need five clear names and no folder tree:

01-Deck-CompanyName-2026-10.pdf
02-Financial-Model-CompanyName-2026-10.xlsx
03-Cap-Table-CompanyName-2026-10.xlsx
04-Founder-Agreements.pdf
05-IP-Assignments.pdf

Date them in the filename. An investor opening a file called Model_final_v3_NEW.xlsx learns something about your company that you did not intend to tell them.

Structured platforms do this better because the fields are named and the versions are current by construction. But a well-named folder of five current documents beats a badly-maintained anything.

The pre-seed test​

Send the link to an advisor who does not know your company. Ask them one question: "Could you write a one-paragraph description of what this company does, what it is raising and what the money is for, using only this?"

If they cannot, no investor will either.

Frequently Asked Questions​

Yes, but a small one. Five documents, current and complete, is the working standard.

Deck, financial model, cap table, founder agreements with vesting, and IP assignment. Beyond that, requests are company-specific.

Most early-stage investors will not sign one and asking can cost you the meeting. Use tiered access instead: share what is safe broadly, and hold financials and cap table for investors who have taken a call.

Unassigned IP, followed by a cap table that does not reconcile to the SAFEs sitting next to it.

A focused day for these five documents, if your paperwork exists. If it takes longer, the delay is telling you about a gap in the paperwork, not about the data room.

The Startup Data Room, Field by Field: What Investors Actually Open in 2026Data Room Red Flags: The 11 Gaps That Stall Due DiligenceData Room Access Tiers: What to Share at First Contact vs Due DiligenceThe Fundraising Readiness Score: How Investors Rate You Before the First Call

RaiseTalks gives you one 143-field record: fill the pre-seed essentials now and complete the rest as your round progresses, so you never rebuild. Start your 30-day trial.