Data room red flags that can kill deals before the first call

Most rounds don't die in the IC meeting. They die in the first hour an analyst spends in the data room. Quietly, in the margin of a spreadsheet, before the partner has even opened the deck. The signal is rarely a single fatal number. It is a pattern of small inconsistencies that tell an experienced eye the company doesn't fully understand its own business.

These are the patterns that end the conversation early.

1. Definitions that drift

The chart says ARR. The footnote says "annualized monthly billings." The board deck says MRR × 12. Three numbers, three definitions, none reconciled.

The moment an analyst has to reconstruct what a metric means, the diligence has stopped being about the company and started being about the data. That is a path no founder wants the conversation to take.

2. Cohorts that don't reconcile to revenue

The cohort table should either reconcile to recognized revenue or include a clear bridge explaining why it does not. Common reasons it doesn't: cohorts built on billings while revenue is recognized ratably, one-time services bundled in with subscription, customers reassigned to a different start month after a contract change. Each is defensible. But only if it's surfaced.

When the gap is unexplained, the underwriter stops trusting either number. The bridge is mechanical to write. The damage from skipping it is not.

3. A churn rate without a definition

Gross or net. Logo or revenue. Monthly or annualized. Trailing-three or trailing-twelve. There are many different numbers a company can call "churn," and they can differ materially on the same underlying data.

A churn figure presented without an explicit definition gets read as the founder picking the most flattering one. Sometimes that's true. Sometimes it isn't. The result is the same.

4. The P&L doesn't match the dashboard

The data room cover page shows $14M ARR. The financial statements show $11.2M of trailing revenue. There may be a perfectly reasonable bridge. Exit-rate vs. recognized, billings timing, deferred revenue. But if that bridge isn't documented, the partner is doing it in their head, and they will assume the gap is unfavorable.

A one-line reconciliation between management metrics and reported revenue removes this entirely.

5. One-off events presented as recurring

A six-figure professional services engagement booked into MRR. A renewal pre-paid in cash and treated as new revenue. A pricing migration that lifted ARR overnight, presented as organic growth.

These are easy to spot in a transaction file. When an underwriter spots them and the founder hasn't flagged them, the read is not "they made a mistake." The read is "what else?"

6. Stale or partial data

The most recent month in the data room is from a quarter ago. The headcount file lists 87 people; the payroll export shows 64. The pipeline export is from before a re-segmentation that the CRO mentioned on the intro call.

Every stale file is an unanswered question. Unanswered questions compound. Three is enough to slow the process. Five is enough to end it.

What this means for the founder

None of these red flags are signs of a bad company. Most are signs of a fast-growing one whose finance function hasn't caught up with its commercial reality. The cost is paid in the data room anyway.

Levian reads your P&L, balance sheet, and transaction files before an investor does. It surfaces the inconsistencies, missing bridges, and metric-quality issues that tend to slow or end the conversation. So when you walk in, the diligence is about the business.