Proportionate Diligence: A Pre-Seed Checklist That Is Not a Series A Checklist
Pre-seed diligence has two failure modes and most investors pick one. Either a Series A process is applied to a $250K cheque, burning weeks on a company with no revenue to analyse, or there is effectively no verification at all beyond a good conversation.
The correct answer is proportionate: a short, specific list, completed in days, that catches the things which are genuinely fatal and skips the things which cannot yet be known.
Key takeaways
- Four items are never optional: entity, cap table reconciliation, IP assignment, founder background verification.
- Everything else is proportionate to cheque size and thesis risk. Founder vesting is worth confirming alongside the cap table.
- Verify the one metric your thesis depends on. Not all of them. The one.
- Two or three customer conversations are worth more than any amount of document review at this stage.
- Target: days, not weeks. If it takes longer, the company's record is the bottleneck, not your process.
The principle
At pre-seed there is very little to diligence. Usually no revenue history, often no product in market, rarely a contract worth reading. You are underwriting a team and a thesis.
So the job is not to verify performance. It is to confirm the company is what it says it is, and that nothing structural will blow up later. Those are different questions and the second one is answerable quickly.
The list below is sorted accordingly: the four non-negotiables first, then the proportionate layer, then what to skip.
The four non-negotiables
1. The entity exists and is in good standing
Incorporation documents, jurisdiction, current standing, registered directors. Five minutes via a registry search.
Sounds trivial. Is occasionally not: dissolved entities, companies incorporated in a jurisdiction nobody mentioned, and cap tables that describe a different legal entity from the one signing.
2. The cap table reconciles
Every shareholder, every SAFE, every note, every option grant, with caps and discounts. Add your proposed investment and check the resulting ownership matches what the founder told you.
The common failures: an unrecorded advisor grant, a SAFE from a friend that nobody logged, an option pool that does not add up.
This is not a trust exercise. Founders rarely misrepresent their cap tables deliberately. They frequently lose track of them.
3. IP is assigned to the company
The most common gap in early-stage diligence and the most expensive to fix late.
Signed assignments from every founder, employee and contractor who contributed to the product. Ask specifically about contractors and about anyone who has left. A codebase with a commit history containing a name that appears nowhere in the paperwork is the classic finding.
If it is not fixed, the company does not own what you are buying into.
4. Founder backgrounds verified
Not read. Verified. Two calls to people who worked with them, and a check that the claimed roles and dates hold up.
At pre-seed the team is most of the investment. Spending an hour here is the highest-return hour in the entire process, and it is the item most often skipped because it feels awkward.
The proportionate layer
Scale these to your cheque size and to what your thesis actually depends on.
| Item | Do it when | Time |
|---|---|---|
| 2-3 customer or user conversations | Almost always. Highest signal per hour available. | 2 hours |
| Verify the one metric the thesis rests on | Always. Pick one, check it at source. | 1 hour |
| Technical review | Deep tech, or if the technical claim is the moat | 2-4 hours |
| Competitive check | Always, briefly. Confirm they know their landscape. | 1 hour |
| Regulatory position | Fintech, health, anything licensed | Varies, can be material |
| Founder reference from a prior investor | If they have raised before. Always revealing. | 30 min |
| Prior round documents | If there is a prior round | 30 min |
| Employment agreements | Larger cheques only | 1 hour |
On verifying metrics: the instinct is to check everything, which is slow and low-yield. Instead, identify the single number your thesis depends on and verify that one at source. If the thesis is "they have found a repeatable channel", verify the channel data. If it is "retention is exceptional", get the cohort export. One verified number beats ten reported ones.
What to skip at pre-seed
- Full financial audit. There is nothing to audit.
- Detailed revenue quality analysis. On $8K MRR this is theatre.
- Market research commissioning. Your own read is worth more than a report at this stage.
- Extensive legal review of customer contracts. Usually there are two, and they are templates.
- Background checks beyond reference calls. Unless something specific prompts it.
- Technical deep-dive on a standard stack. If it is a normal web application, the architecture is not the risk.
Each of these is appropriate at Series A. At pre-seed they consume weeks and return almost nothing, and the cost is not only yours: a founder held in diligence for six weeks by a $200K cheque has lost momentum with everyone else.
The timeline
| Day | Activity |
|---|---|
| 1 | Entity, cap table, IP assignment. Request anything missing. |
| 2-3 | Founder references and background verification |
| 3-4 | Customer or user conversations |
| 4-5 | Metric verification, technical or regulatory review if applicable |
| 5 | Write the memo, make the call |
One week. If it takes three, the bottleneck is usually the company's record rather than your process - which is itself a finding worth recording, because it predicts what working with them will be like.
Where structure changes the arithmetic
Most of week one is spent requesting, waiting for and reassembling information that already exists somewhere. The verification itself is fast; the collection is not.
A company with a complete structured record collapses days 1 and 2 into about an hour, because the entity details, cap table, instruments and assignments are already present and current. What remains is the genuinely irreducible work: talking to people.
This is the same argument as the founder-side one, seen from the other chair. The record is not paperwork. It is the difference between a one-week process and a three-week one, and at pre-seed that difference determines whether the good companies stay in your funnel.
Frequently Asked Questions
About a week. Four non-negotiable checks, a handful of conversations, and one verified metric.
Entity and good standing, cap table reconciliation, IP assignment, founder background verification. Then customer conversations and the one metric your thesis depends on.
The four non-negotiables are, regardless of size. They protect against structural problems that no cheque size makes acceptable.
IP assignment, particularly from contractors and departed contributors.
Five working days. Longer usually indicates the company's records are incomplete.
Read next
The Modern Dealflow Stack: From Inbox Chaos to a Scored Pipeline
The Investment Memo, Standardised: What Belongs in a Pre-Seed IC Memo
Angels and Micro-Funds: Running Real Diligence Without an Analyst Team
Data Room Red Flags: The 11 Gaps That Stall Due DiligenceWhen a founder grants you due-diligence access on RaiseTalks, you get their full 143-field record in one step, so week one is verification instead of collection. Start your 30-day trial.