The finance bar for a seed round is low enough that most founders clear it without thinking about it. That is exactly why the Series A version catches people out: nothing about the seed process teaches you what the next one requires, and the gap between them is larger than the gap between seed and pre-seed.
At seed, investors are underwriting a team and a market with barely any operating history to examine. At Series A, someone reads your books, and the questions move from what you are building to whether your numbers hold together.
What actually changes
| Seed | Series A | |
|---|---|---|
| What is being underwritten | Team, market, early signal | Unit economics and repeatability |
| Financial diligence | Light, often days | Detailed, typically weeks |
| Accounting basis | Cash-basis usually tolerated | Accrual expected |
| Metrics | Directionally right is enough | Must reconcile to the statements |
| Model | A plan | Stress-tested, reconciled to actuals |
| Cap table | Straightforward | Every SAFE, option and 409A reflected |
| Data room | A folder | Organised, complete, consistent |
The row that causes the most pain is the accounting basis. Cash-basis books that were fine at seed have to become accrual, and doing that conversion during a live round means restating history at exactly the moment consistency is being examined.
The five things a Series A investor checks
- Accrual, GAAP-ready statements with a revenue recognition policy applied consistently all year, not adopted the month before the raise.
- A model that reconciles to your actuals. Divergence between the two is read as either poor control or optimism, and neither helps.
- Metrics that tie back to the statements. If ARR in the deck cannot be reconciled to revenue in the P&L, the deck stops being evidence.
- A clean cap table, with SAFEs converted correctly, the option pool reflected, and 409A work current.
- Consistency between what you told existing investors through the year and what you are telling a new one now.
The fifth is the quiet one. Investors talk to each other and read your old updates. A number that moved without explanation between a board update and the pitch is a question you will have to answer, and it is far better to explain it yourself first.
What does not change
Series A diligence is a review exercise, not an audit. Investors and their advisors read the books closely and expect answers to tie out, but they are not issuing an opinion on your financial statements. A formal audit is rare at this stage and generally only appears later, when a lender or an acquirer requires one.
So audit-readiness at Series A means being able to answer where a number came from quickly, from records you already keep. It does not mean commissioning an audit, and a provider suggesting you need one to raise a Series A is usually selling one.
The cheapest version of this is to run accrual from incorporation. Nearly every expensive Series A finance problem traces back to a cash-basis history that has to be converted under time pressure, with the conversion itself becoming a diligence topic.
The timeline that works
Six months before you open a round is when the work is cheap. One month before, it is expensive and visible.
| When | What should be true |
|---|---|
| 6 months out | Books current and accrual; revenue policy settled |
| 4 months out | Model reconciles to actuals; metric definitions written down |
| 3 months out | Data room assembled; cap table and 409A confirmed current |
| 1 month out | The awkward questions answered before an investor asks them |
If you start at month three with books three weeks behind, the first month goes on cleanup rather than preparation. That is the same mistake as hiring a CFO to fix bookkeeping, and it costs the same thing: time you needed for the actual raise.
The questions worth rehearsing
Most of the finance questions in a Series A process are predictable, which means the difference between a smooth process and a painful one is preparation rather than luck.
- Walk me through how you recognise revenue, and show me a contract that demonstrates it.
- Why does this metric differ from the one in your Q2 board update? Have the answer before anyone finds the discrepancy.
- What is in cost of revenue, and what did you decide to leave out?
- How did you calculate burn, and is the method the same in every month shown?
- What are the three assumptions the model is most sensitive to?
The last one is where founders most often lose ground, because answering it well requires having stress-tested your own model rather than only having built it.
The data room, itemised
Assemble the data room is advice that assumes you know what goes in it. For the finance section of a Series A, this is the list. Most of it should already exist if the books are current, which is the point.
- Monthly financial statements for every month since inception, or at least the last two full years: P&L, balance sheet and cash flow, on a consistent basis throughout.
- A trial balance as at your most recent close, and the general ledger detail behind it.
- The revenue recognition policy, written down, with one or two contracts that demonstrate it being applied.
- Your operating model, with actuals through the most recent closed month and the link between the two visible rather than asserted.
- A metrics file: ARR, retention, gross margin, burn and runway, each with its definition and each traceable to the statements.
- Cap table with all instruments converted, the option pool shown, and the most recent 409A valuation.
- Payroll summary by function, current headcount, and any signed offers not yet started.
- Tax filings: federal and state returns filed to date, Delaware franchise tax, and any R&D credit claimed. See the R&D tax credit for startups.
- Material contracts: your largest customers, any contract with unusual terms, and anything with a change-of-control clause.
The single most useful thing you can do with this list is assemble it six months early and then keep it current, rather than building it during a live process. A data room built under time pressure is where inconsistencies get introduced, not where they get caught.
Five definitions to pin down before anyone asks
Write the metric definitions down is the item founders most often skip, because the metrics feel self-evident until someone asks a precise question about them. These five are where the ambiguity actually lives, and being unable to answer crisply reads as not knowing your own business.
- ARR. Contracted recurring revenue, or last month multiplied by twelve? The second includes anything one-off that happened to land in that month, and the difference between the two is the first thing a diligence analyst will recompute.
- Churn. Logo churn and revenue churn tell different stories, and a business losing small customers while growing large ones looks healthy on one and poor on the other. State which you are quoting, and be ready with the other. See net revenue retention.
- Gross margin. What is in cost of revenue? Hosting is uncontroversial. Support, customer success, and the engineering time spent on delivery rather than product are all judgment calls, and your answer needs to be the same one you used last quarter.
- Burn. Gross or net, and does it include or exclude one-off items? A number that quietly switches definition between board updates is worse than a bad number consistently reported. See what is a good burn rate.
- Customer. Per contract, per logo, or per billing account? This sounds trivial until it changes your per-customer economics by a wide margin because one parent company holds six contracts.
Write these five definitions in a single document and put it at the top of the metrics section of your data room. It takes an hour, it forces you to resolve the ambiguities before someone else finds them, and it signals control more effectively than any individual number does.
Where Zinance fits
Because books close daily and run on accrual from the start, the diligence-ready state is the normal state rather than a project you begin before a raise. Bookkeeping, tax and R&D credits sit with one team, so the statements, the model and the return are built from one set of numbers when someone starts checking. See the Series A audit-readiness checklist and do you need a CFO before your Series A.