How to Systematically Estimate the Chance a Startup Will Survive to Exit

How to Systematically Estimate the Chance a Startup Will Survive to Exit

The standard survival rate literature will tell you 50% of startups make it to year five. That's not the number a GP needs. The number that matters is: what is the probability this specific startup survives to an exit above $1M — and what are the eight factors that determine it?

Framework Pv × Pc · Exit Likelihood Scoring

Modules 2 & 3

Outcome Probability of Exit Above $1M

Why "50% of startups survive 5 years" is the wrong number for a GP

When researchers measure startup survival, they mean "still operating." When a GP needs to assess survival, they mean something entirely different: does this company reach an exit valued above $1M? Those are different questions — and the industry has no systematic method for answering the second one. Most firms fall back on intuition. Different partners score the same deal differently. The same partner scores deals differently on different days. There is no baseline, no audit trail, and no way to detect whether assessment quality is drifting.

This is not a minor inconvenience. Systematic bias in exit likelihood assessment compounds over time. A firm that consistently overestimates the chance of a successful exit for deals sourced by a particular partner — or underestimates commercial risk in familiar verticals — carries that bias invisibly across every investment decision until the DPI doesn't show up.

The Pv × Pc framework makes exit likelihood assessment explicit, comparable, and auditable — producing a single number that can be defended, documented, and updated as new information arrives.

Survival defined as exit above $1M

Survival in this framework means one thing: the company reaches an exit valued above $1M. Ps is the probability of that outcome — not the probability of staying alive, not the probability of raising another round. Ps = Pv × Pc: Venture Viability multiplied by Commercial Scaling probability. These are independent dimensions — a technically sound product does not guarantee a commercial engine, and vice versa. Both must be scored. Neither substitutes for the other.

Pv — can this company survive long enough to exit?

Pv scores the four internal factors that determine whether the company has the foundations to reach an exit at all: Team, Product, Traction, and Timing. These are scored independently and combined into a single Pv score that answers one question: can this company get to the starting line?

The most important feature of the Pv model is the $1M ARR threshold. Once a company crosses $1M in annual recurring revenue, Pv is treated as 1.0 — it has already demonstrated it can generate revenue at scale. The survival question shifts entirely to Pc: can it build a commercial engine large enough to generate an exit above $1M?

Pc — can this company build an engine large enough to exit?

Pc scores the four external factors that determine whether the company can build a commercial engine capable of generating an exit above $1M: Market size, Distribution quality, Differentiation defensibility, and Competitive intensity. A company with a 0.9 Pv and a 0.3 Pc has a Ps of 0.27 — a 27% chance of the exit outcome a GP needs. Most investment thresholds sit considerably higher than that.

What the Ps score tells you beyond a single number

The Ps score is not just a probability — it is a structured diagnostic that tells you where the risk to exit lives. A company with high Pv and low Pc has a commercial scaling problem: the product works but the engine to monetise it at exit scale doesn't exist yet. A company with low Pv and high Pc has the opposite — a large addressable market but insufficient internal foundations to reach it.

That diagnostic drives different structuring decisions. A low-Pv company might warrant a smaller initial check with a defined ARR milestone that re-triggers full evaluation. A low-Pc company might warrant a condition on distribution channel validation before the investment closes. Without the framework, both look like "risky deals with upside." With it, the specific obstacle between this company and a meaningful exit is explicit — and addressable.

Two dimensions. Eight factors.
One exit likelihood score.

The Pv dimension scores four internal factors that determine whether the company as a venture can survive. The Pc dimension scores four external factors that determine whether it can build a commercial engine. Each factor is scored independently on a calibrated scale — the specific rubric, weighting, and combination mechanics are what make the system repeatable and auditable across a firm.

Pv — Venture Viability

Four factors assess whether the company has the internal foundations to survive as a venture. Once a company crosses a defined revenue threshold, Pv is treated as proven — the risk shifts entirely to commercial scaling.

Factor details covered in Module 2

Pc — Commercial Scaling

Four factors assess whether the company can build a durable commercial engine. These are assessed independently of Pv — technical excellence does not guarantee commercial success, and the two dimensions must be scored without reference to each other.

Factor details covered in Module 3

Ps maps directly to
an investment decision.

The output of Pv × Pc is a single probability of survival to exit. That number maps to a tiered decision framework — invest, watch with conditions, or pass — with each tier carrying a defined action and documentation requirement. The threshold values are set at the fund level and applied consistently across all deals.

The specific threshold ranges, what each tier requires from the GP, and how to set thresholds appropriate to your fund's risk tolerance are covered in Module 3 of the VC Risk curriculum.

The key principle

Thresholds are set once, pre-deal, at the fund level — not negotiated deal-by-deal. This is what makes the system auditable and bias-resistant. A GP who adjusts the threshold for a deal they love has broken the system.

How to score the chance any startup survives to exit — step by step

Score Pv — the four internal factors that determine whether the company can reach an exit

Score Team, Product, Traction, and Timing each on a 0–1 scale using the calibrated rubric. Do not average — combine multiplicatively after applying factor weights. If Traction includes $1M+ ARR, set Pv = 1.0 directly.

Pv = f(Team, Product, Traction, Timing)

Score Pc — four factors independently

Score Market, Distribution, Differentiation, and Competition (inversely) on the same 0–1 scale. Pc is independent of Pv — assess it without reference to the Pv scores you just assigned.

Pc = f(Market, Distribution, Differentiation, Competition⁻¹)

Compute Ps — the probability of exit above $1M

Multiply Pv × Pc to get Ps — the probability this company exits above $1M. Then compute expected value: Ps × P50 exit value from the cohort distribution. That EV is the single number that enters the deal ranking system and drives the investment decision.

EV = Ps × P50 exit value

Compare against the investment threshold

Apply the fund's pre-set Ps threshold. If Ps exceeds the threshold, proceed. If it falls in the watch range, define the specific milestone that would move it above threshold. If below — pass and document.

Decision = Ps vs. fund threshold

Re-score at every financing stage as new information updates the exit likelihood

Ps is not a one-time assessment. Re-score at every follow-on decision. As new information arrives — ARR milestones, competitive moves, team changes — the Pv and Pc scores update. This is what makes the framework a dynamic management tool, not just a due diligence checklist.

Ps(t+1) = Pv(t+1) × Pc(t+1)

How to Systematically Estimate the Chance a Startup Will Survive to Exit

Why "50% of startups survive 5 years" is the wrong number for a GP

Pv — can this company survive long enough to exit?

Pc — can this company build an engine large enough to exit?

What the Ps score tells you beyond a single number

Two dimensions. Eight factors.
One exit likelihood score.

Ps maps directly to
an investment decision.

How to score the chance any startup survives to exit — step by step

Five frameworks.
One complete system.

Score the exit likelihood of every deal you see.

How to Systematically Estimate the Chance a Startup Will Survive to Exit

Why "50% of startups survive 5 years" is the wrong number for a GP

Pv — can this company survive long enough to exit?

Pc — can this company build an engine large enough to exit?

What the Ps score tells you beyond a single number

Two dimensions. Eight factors.One exit likelihood score.

Ps maps directly toan investment decision.

How to score the chance any startup survives to exit — step by step

Five frameworks.One complete system.

Score the exit likelihood of every deal you see.

Two dimensions. Eight factors.
One exit likelihood score.

Ps maps directly to
an investment decision.

Five frameworks.
One complete system.