The Risk Matrix Lies to You. Write the Autopsy Instead.

Every diligence deck has one. A tidy grid, red-amber-green, risks plotted on two axes: likelihood and impact. It looks like rigor. It photographs well in an investment committee memo. And most of the time, it is theater.

Here is the uncomfortable part. By the time you build the matrix, you have already decided. The founder is charismatic, the market is hot, the round is competitive, and the grid exists to ratify a “yes” you reached three meetings ago. So the scary risks quietly migrate toward the lower-left corner. Low likelihood. Manageable impact. Green.

A risk matrix does not surface what will kill the company. It launders the conviction you walked in with.

What the matrix actually does wrong

The color-coded grid has three structural flaws, and none of them are about the colors.

First, it treats risks as independent. You score “key-person dependency” and “slow enterprise sales cycle” as separate cells. But companies do not die of one thing. They die of a chain: the technical co-founder leaves, which slows the roadmap, which blows the enterprise pilot, which kills the Series A, which triggers the down round nobody survives. The matrix has no column for causality.

Second, it rewards vagueness. “Market risk: medium.” Medium what? Over what horizon? Medium is the score you give something you have not actually thought about. Precision is expensive, so the matrix fills with hedges that feel responsible and commit to nothing.

Third, and worst: it is motivated reasoning with a coordinate system. You know the answer you want. The grid lets you place each risk exactly where it needs to sit to preserve that answer, and the format makes it look neutral.

A risk matrix tells you how you feel about a deal. An autopsy tells you how the deal dies.

Write the autopsy instead

Flip the exercise. Do not ask “what could go wrong and how bad would it be.” Assume it already went wrong. It is three years from now, the company is dead or a zombie, and you are writing the postmortem. Your only job is to explain, in specific and slightly embarrassing detail, why.

This is the pre-mortem, borrowed from Gary Klein and made famous by Kahneman, and it works because of a documented quirk in how brains reason. “Prospective hindsight” — imagining an outcome as already fixed and generating causes for it — produces roughly 30% more, and more concrete, reasons than open-ended forecasting. Certainty unlocks specificity. When failure is a given, you stop defending the deal and start explaining it.

The framework: seven steps to a real autopsy

1. Fix the date and the verdict. “It is Q3 2029. This company has failed.” Not “might fail.” Has. Write it in past tense from the start.
2. Generate the death certificate, alone first. Everyone on the deal team writes their own cause of death before anyone speaks. Groupthink is what the matrix protects; solo drafting breaks it.
3. Cluster into distinct failure modes. You are looking for genuinely different stories, not five flavors of “ran out of money.” Money is how companies die; it is never why.
4. Build the collapse timeline. For each mode, walk it backward quarter by quarter. What broke in Q8? What was the tell in Q4? What was already visible at close?
5. Extract the foreshadowing. This is the payoff. For each timeline, name the signal that was observable during diligence — the flag you could have caught but were primed to explain away.
6. Convert flags into diligence questions. Every red flag becomes a thing you go verify before wiring money, not a cell you shade amber.
7. Then, and only then, size it. Now build your matrix if you want one. It will be honest, because it is downstream of the stories instead of upstream of your conclusion.

A worked example

Say you are looking at a Series A in a vertical AI SaaS company. $40M pre, strong logos in the pilot pipeline, a technical founder who demos beautifully. The matrix says: team green, market amber, tech green, competition amber. Clean deal.

Now the autopsy. It is 2029. The company is being acqui-hired for less than it raised. Cause of death:

• Q1–Q2: Two marquee pilots convert, but each demands heavy custom integration. The team says yes to everything to hit the logo count.
• Q3–Q4: The product bifurcates into two bespoke deployments. Roadmap velocity halves. The founder is now in the weeds of customer #1’s data pipeline instead of shipping.
• Q5–Q6: A horizontal incumbent ships a “good enough” version of the core feature as a checkbox. New logo growth stalls. The moat was a feature, not a product.
• Q7–Q8: Burn is high, growth is flat, the Series B market wants efficiency. No inside lead. The party ends.

The foreshadowing, visible at close: the pilots were services revenue in a SaaS wrapper. That was checkable in diligence — pull the implementation hours per customer, ask what percentage of “product” was actually solutions engineering. The matrix scored “market: amber” and moved on. The autopsy handed you one blunt question that would have repriced the whole deal.

Common mistakes

• Hedging the verdict. “It struggled.” No. It died. Past tense, no qualifiers, or the mechanism kicks in only halfway.
• Stopping at “ran out of runway.” That is a symptom. Keep asking why until you hit something you could have checked.
• Writing failures you are already comfortable with. The useful autopsy names the risk you have been rationalizing, not the one on the founder’s own slide.
• Never mapping flags back to diligence. An autopsy that does not change what you go verify is just creative writing.

The problem

Done properly, this is slow and genuinely hard. Five distinct failure narratives, each traced backward across eight quarters, each mined for its diligence tell — that is a day of focused, imaginative, adversarial thinking per deal. It fights your own optimism, which is exhausting. It requires you to be creative about your own blind spots, which is close to a contradiction.

So it gets faked. The autopsy becomes a bullet point. The pre-mortem becomes “risks” on slide 14. And the tidy grid comes back, because a grid takes twenty minutes and looks just as authoritative in the committee room. The rigor gets compressed into color exactly because the real version is too much work to do by hand under deal pressure.

Risk Matrix

This is the specific gap Risk Matrix, in the VentureVerse suite, is built to close. You give it the company. It assumes the company already failed, three years out, and writes the autopsy for you.

Concretely, it produces three things: five distinct failure scenarios — not five versions of the same death — a quarter-by-quarter collapse timeline for how each one unfolds, and the diligence red flags that foreshadowed it, mapped back to what was observable before you invested. It does the part humans skip: the disciplined, unflattering, backward walk from a fixed failure. You bring judgment; it removes the excuse that the honest version takes too long.

Try Risk Matrix →

It’s not the only way

Risk Matrix is one tool, and it has a real trade-off. Here is the honest landscape.

The bottom line

The risk matrix is not useless, but it belongs at the end, not the start — as a way to communicate a conclusion, never to reach one. If you want a true read on a deal, refuse the empty grid and write the death certificate first. Assume the failure, explain it in specifics, and drag every explanation back to a flag you could have caught. Do that and the “yes” you eventually give is one you have actually earned, not one a coordinate system rubber-stamped for you.

Try Risk Matrix · Get The Brief