Red Flags in Startup Due Diligence: A Field Guide

A red flag is a signal serious enough that it should pull a candidate's score down on purpose, rather than sit politely in a list where a strong average can bury it. The danger in due diligence is not missing information, it is finding a real problem and letting it get averaged away. This guide covers the red flags worth watching, how to tell a genuine one from ordinary risk, and how to handle it without tipping into paranoia.

Every serious problem a program funds by accident was usually visible beforehand, sitting quietly among the ordinary risks, outweighed by everything the company did well. That is the specific failure red flags exist to prevent. A strong company with one disqualifying issue and a mediocre company with none can post the same overall score, and a process that treats a governance problem like a slightly weak market is a process that will eventually fund the wrong thing.

This guide walks through the categories of red flag worth knowing, from claim inconsistencies to governance to cap-table problems, how to distinguish a real red flag from routine risk, and the discipline that keeps flagging useful rather than a reflex that kills every deal.

What a red flag actually is

A red flag is not just a negative. Every company has negatives, and treating each one as disqualifying means never funding anything. A red flag is a negative serious enough that it changes the decision on its own, regardless of how strong the rest of the picture is. The defining test: would you still commit if this were true and everything else were excellent? If the honest answer is no, it is a red flag. If the answer is "it depends on the rest," it is ordinary risk to be weighed, not flagged.

The reason the distinction matters is arithmetic. In any scoring, a single strong-average number lets a serious problem hide behind strength elsewhere. A red flag is the mechanism that pulls the overall score down on purpose, so the problem stays visible to the committee instead of being smoothed into an acceptable total. Flagging is not about being negative. It is about refusing to let a real deal-breaker be averaged away.

Claim and consistency red flags

The first place red flags show up is in the claims themselves, when what the company says does not survive checking.

• A contradicted material claim. Not a rounding difference, but a headline claim that public evidence contradicts: a partnership that the named partner does not confirm, a metric that authoritative sources refute. A company willing to state something contradicted has told you how it will report to you later.
• Numbers that do not reconcile. Traction, revenue, and headcount that cannot all be true at once. Internal inconsistency is a signal even when you cannot say which figure is wrong.
• Evasion under verification. A founder who cannot or will not substantiate a claim they made, or who reframes it each time it is questioned. The claim was checkable and they avoided the check.

The honesty rule here matters as much as the flag: a metric that legitimately shifted over time, a customer count from three years ago against a current source, is not a contradiction. Check the date before you flag it. A red flag is a real inconsistency, not a stale number.

Governance and people red flags

The second category is about who runs the company and how, and it is often the most predictive.

• Undisclosed conflicts or history. A material fact about a founder or the company that was not disclosed and should have been. The problem is less the fact than the non-disclosure.
• A pattern, not an incident. One departed cofounder is life; a revolving door of early leadership is a signal. Distinguish a single event from a pattern.
• Governance that concentrates risk. Structures that leave no check on a single decision-maker, or that make the company hard to hold accountable. This is not about founder control per se, it is about whether anything catches a bad call.

Structural and cap-table red flags

The third category is about the shape of the company, the parts that are hard to fix after the fact.

• A broken cap table. Ownership so fragmented, or so concentrated in inactive early parties, that future rounds are compromised. A cap table is hard to unwind, so a problem here is durable.
• Unresolved legal or IP exposure. Ownership of the core technology in question, unresolved disputes, or licensed data being presented as an owned moat. These do not improve on their own.
• Dependence with no fallback. A single point of failure the company cannot survive losing: one customer, one platform, one provider whose terms could change. Concentration itself is the flag.

How to handle a red flag without paranoia

The failure mode on the other side is treating everything as a red flag until no company survives. A discipline keeps flagging useful. First, apply the test consistently: would this alone change the decision. If not, it is risk to weigh, not a flag. Second, verify before you flag: a suspected contradiction that turns out to be a stale metric or a misread source is not a red flag, and flagging it erodes trust in the ones that are real. Third, record the flag with its evidence, so the committee weighs the actual problem rather than a vague unease, and so the decision, either way, is defensible.

A red flag is not an automatic no. It is a signal loud enough that the decision must consciously account for it. Sometimes a committee looks at a real red flag and proceeds anyway, with eyes open. That is a legitimate outcome. What is not legitimate is never seeing the flag because a strong average hid it. The goal is not to kill deals. It is to make sure the serious problem is on the table when the decision is made.

Frequently asked questions

What is a red flag in due diligence? A negative signal serious enough to change the decision on its own, regardless of how strong the rest of the case is. It should pull the overall score down on purpose so it stays visible, rather than being averaged into an acceptable total where a strong picture hides it.

How is a red flag different from ordinary risk? By one test: would you still commit if this were true and everything else were excellent? If no, it is a red flag. If "it depends on the rest," it is ordinary risk to weigh. Every company has risks; a red flag is the subset that decides on its own.

What are the most common startup red flags? Contradicted material claims and numbers that do not reconcile, undisclosed conflicts or a pattern of leadership churn, governance that leaves no check on a decision, a broken or fragile cap table, unresolved IP or legal exposure, and critical dependence on a single customer, platform, or provider.

Does a red flag automatically mean rejecting the company? No. A red flag means the decision must consciously account for it. A committee can look at a real flag and proceed with eyes open, and that is legitimate. What is not legitimate is failing to see it because a strong average buried it. The flag forces the problem onto the table.

How do I avoid flagging everything? Verify before flagging, and apply the "would this alone change the decision" test consistently. A suspected contradiction that turns out to be a stale metric is not a red flag, and over-flagging erodes trust in the real ones. Record each flag with its evidence so it is weighed as a specific problem.

The bottom line

A red flag is a problem serious enough to decide on its own, and the job of due diligence is to make sure it is never averaged away behind everything a company does well. Watch the claim, governance, and structural categories, verify before you flag, apply the "would this alone change the decision" test, and record each flag with its evidence. A red flag is not an automatic no. It is a guarantee that the serious problem is visible when the decision is made.

How this shows up in Deckwise

In Deckwise, a red flag is a first-class mechanism, not a note in the margin. When a material claim is contradicted or a serious negative surfaces during verification and due diligence, it pulls the candidate's overall score down on purpose, so it cannot hide behind strength elsewhere, and it appears in the memo with its evidence for the committee to weigh. Verification runs registers first, then sourced web search, and a metric that shifted over time is qualified rather than flagged as a lie. The flag makes the problem visible; the human still decides what to do about it.

Related: The Complete Guide to Startup Due Diligence · How to Write an Investment Memo · The Deckwise Method