Every founder believes the data room is read in order — executive summary, pitch deck, financials, legal. It is not. There is a pattern, and knowing it changes what you prepare and in what order.
The actual reading sequence
First: the cap table. Before an investor reads a single line of your model, they want to know the structure. Who owns what. What the option pool looks like. Whether there are side letters, participation rights, or pro-rata provisions that would complicate the round. The cap table tells them whether a deal is structurally possible before they decide whether the business is worth pursuing.
A messy cap table — missing option grants, stale 409A valuations, unauthorized share classes — kills deals before the model is opened. Clean it before the fundraise starts, not during.
Second: the model. Not the deck. The model. Investors who are serious about diligence skip the pitch deck on the first pass — they've already heard the story in the meeting. The model is where they test the story against math.
They are not reading every row. They are checking three things: Does the model close? Are the assumptions labeled and sourced? And is there a scenario where this business works — not the base case, but the model that answers "what has to be true for this to be a good investment."
Third: the gap. This is the one founders don't expect. After reading the cap table and the model independently, investors look for the gap between them. Specifically: does the fully diluted share count in the model match the cap table? Does the amount raised in the financing schedule match what the cap table shows was actually issued? Do the option grants in the model match the option pool in the cap table?
The gap, when it exists, is usually innocent — a model built before the last extension, an option grant that was approved but not yet in the data room. But it creates a question that has to be answered, and questions slow deals.
What this means for preparation
Prepare the cap table before the model. The cap table should be final — fully updated, 409A current, all grants documented — before you share the model with anyone. The model should reference the cap table, not the other way around.
Label every assumption. The investors reading your model have seen hundreds of them. They know when a churn rate is unsupported and when a growth assumption is disconnected from the sales capacity model. A labeled assumption — "2.3% monthly churn, based on cohort data from Q3 2025 — three cohorts" — invites a question. An unlabeled assumption invites suspicion.
Build the gap-check yourself. Before you share the data room, reconcile the model to the cap table manually. Fully diluted shares, proceeds by security class, option pool size. If you find a gap, you have time to close it. If the investor finds it first, you spend two weeks on a question that should have been answered before the process started.
The question investors are actually asking
Every data room exercise is trying to answer one question: Is there a management team that understands the business well enough to be trusted with our capital?
The model is not a forecast. It is a test of whether you understand the mechanics of your own company — how revenue converts to cash, how headcount drives cost, how the balance sheet reflects the operating reality. Investors who find a model that closes, with labeled assumptions, reconciled to a clean cap table, read it differently than one that doesn't. Not because the business is better, but because the team is demonstrably better.