What goes in the data room before the second meeting.

Most founders open the data room too late.

The reflex is to wait — open the room when the term sheet is being drafted, or when the partner explicitly asks. The thinking is that handing over diligence materials early is somehow giving up leverage, or showing too much, or rushing.

The data we see on rounds that close fastest says the opposite. The founders who get to a term sheet most quickly open a scoped, well-prepared data room after the first meeting — that is, before the second one. Not because the partner asked. Because the partner was always going to ask, and the founders who anticipated it saved the round a week.

This piece is the seed-stage data room packet. What is in it, why each section earns its place, and what scoping per viewer looks like.

The seed packet, in six sections.

The seed-stage standard data room has 38 items across six sections. That number is an approximation — some companies need fewer, some need more — but the section structure is stable.

• Corporate. 6 items.
• Financials. 8 items.
• Product. 5 items.
• Market. 4 items.
• Team. 6 items.
• Legal. 9 items.

Below, each section in detail, with the items most founders forget called out.

Corporate (6 items).

The basic legal artifacts that prove your company exists and is yours.

1. Certificate of Incorporation (or the equivalent in your jurisdiction).
2. Bylaws.
3. Board consent and minutes from prior meetings.
4. Cap table snapshot (current, with full SAFE waterfall).
5. 409A valuation snapshot, if you have one. Many seed companies do not. If you do, include it. If you do not, include a short note saying so — its absence is fine, but unexplained absence reads worse than explained absence.
6. Founders' agreements (vesting schedules, IP assignment).

The two most common omissions: the board consent for the last raise, and the IP assignment from each founder. Both are easy to find, both signal sloppiness when missing.

Financials (8 items).

The block that gets the most scrutiny on a seed diligence.

1. Trailing 12-month income statement.
2. Trailing 12-month cash flow.
3. Current balance sheet.
4. 3-year forecast. The single most-read document in the data room. Make it a clean model — assumptions sheet, outputs sheet, monthly cohorts.
5. Bookings vs. revenue reconciliation, if relevant.
6. Burn and runway snapshot.
7. KPI dashboard (MRR, CAC, payback, gross margin, NDR).
8. Bank statements for the last 6 months.

A common omission: the assumptions sheet. Partners will assume any forecast without an explicit assumptions sheet was built backwards from a target. They are usually right.

A second common omission: the bank statements. Founders sometimes feel weird sharing them. Investors at a seed almost always want to see them — they are a quick sanity check on the financials, and their absence raises the partner's eyebrow at no benefit to you.

Product (5 items).

The block most founders enjoy assembling and most partners read least carefully.

1. Product overview deck (separate from the pitch deck — longer, more screenshots).
2. Architecture diagram.
3. Roadmap (next 12 months).
4. Engineering team org chart.
5. Security and compliance posture. SOC2 status, GDPR, HIPAA if relevant. If you are pre-SOC2, say so. The partner is not going to disqualify you over it; they will note it.

A surprisingly common omission: the architecture diagram. Many seed founders feel it is not necessary at seed. Partners with technical depth — especially at funds with strong technical partners — appreciate it as a signal of clarity even if they do not read it closely.

Market (4 items).

1. TAM/SAM/SOM analysis.
2. Competitive landscape map.
3. Customer reference list (the people, not the references themselves — references go in a separate scoped folder).
4. Win/loss data, if you have any.

The competitive landscape is where founders most often overreach. The honest version of a competitive landscape is shorter and more credible than the comprehensive version. Three or four direct competitors, named, with a sentence each on how you differ. Skip the seventeen adjacent companies. The partner is not going to be persuaded by a wider map; they will be persuaded by a sharper one.

Team (6 items).

1. Founder bios (longer than the deck version).
2. Senior team bios.
3. Org chart.
4. Customer references — the actual list. Names, titles, contact info, with a note on which references the partner has clearance to contact and when.
5. Two reference calls with prior colleagues of each founder.
6. A list of hires planned in the next 12 months, with target start dates and target compensation.

The customer references are the section most founders forget at seed, and the section most partners notice when missing. By the time of the second meeting, the partner has been thinking about who would vouch for you. Make it easy to do that.

Customer references in a seed data room are not a defensive artifact. They are the most credible single signal you can put in front of an investor.

Legal (9 items).

The block that is most procedural and most useful for showing diligence.

1. All prior financing documents (SAFEs, convertible notes, prior priced rounds if any).
2. Form of SAFE you intend to use this round.
3. Term sheet from prior round, if any.
4. IP assignments from contractors and former employees.
5. Major customer contracts (top 5 or top 10 by revenue).
6. Major vendor contracts.
7. Litigation history (a one-page note saying "none" if there is none).
8. Open source license inventory, if your product depends on it.
9. Privacy policy and terms of service.

The litigation note is often missing. Investors will assume "no note" means you forgot, not that there is no litigation. Include the note even when it is one sentence long.

What scoping per viewer looks like.

The 38-item packet is not the right thing to hand to every investor.

A tier C investor doing first-call diligence does not need to see your bank statements. A tier A partner deep into term sheet conversation does not need to see your form of SAFE — they are about to send their own.

The discipline of scoping is to send a subset of the room to each investor, matched to where they are in the process. The scopes we recommend at seed:

• First-call diligence scope. 12 items. Corporate basics, the deck, the forecast, the KPI dashboard, customer reference list (names only), the form of SAFE. Sent ahead of the second meeting.
• Second-meeting diligence scope. 24 items. Adds bank statements, full financials, architecture diagram, roadmap, security posture, customer reference contact info.
• Term sheet diligence scope. Full 38 items. Once a term sheet is being drafted, the partner sees everything.

In Arx, each scope is a saved view of the data room; you assign a scope to a viewer with one click, and the room renders only those folders for them. Watermarks and expiry are per-viewer. Other tools require manual folder management; the discipline is the same either way.

What about the NDA?

Most seed-stage investors will not sign an NDA. Don't ask them to. The room itself should not contain anything that is truly secret — the IP assignments are about ownership, not invention; the customer contracts are about commercial terms, not trade secrets; the financials are about the company's shape, not its private moats.

If you have something truly confidential — a hardware schematic, a proprietary algorithm — it does not belong in the round-one data room. Save it for a one-off conversation post-term-sheet, under a one-off NDA.

Common mistakes.

A short list of things we see repeatedly:

• Including everything from the start. A 38-item room sent to a tier C investor on first contact reads as "I am unfocused." Send less, sooner; more, later.
• Forgetting the customer references. The most common single omission.
• Treating the forecast as a one-time artifact. The forecast in your data room should match the version you would defend on a call this morning. If it doesn't, update it before you share the room.
• Letting the room go stale. A data room that says "as of March" in May is worse than a current room. Date every artifact. Refresh monthly during a round.
• No watermarks. Watermark every PDF with the viewer's email. The cost is zero. The benefit, on the rare day a deck leaks, is meaningful.

The summary.

Open the room before the second meeting. Use 38 items, in six sections, with three scopes by viewer stage. Include the things founders most commonly forget — customer references, bank statements, the architecture diagram, the litigation note.

A well-prepared room does not win a round. A poorly-prepared room can lose one. Treat the room as table stakes, do the work once, and stop thinking about it for the rest of the round.