Nasr Khan

08 Diligence

Product diligence scorecard

What product risks does this company carry, and what should we ask for before the term sheet?

For investors and investment committees. Takes about five minutes. Opens with an example company.

What you know so far

1The deal
2Thirteen product checks

Answer from the data room, the management meeting or your own checks. Pick the last option when you don't have the data yet.

3Numbers, if you have them

Leave at 0 if you don't have them yet.

%
%

The answer

Product diligence report

Risk
RetentionPlatform dependence

Information request, ready to send

Ahead of our next meeting, could you share the following?

    Take the result with you

    A PDF report with your inputs, results, charts and method. Or the same as plain text for an email.

    How this tool works

    What it does

    It turns what you know about a company's product into a one-page risk report. Then it writes the information request to send the founders: data where you have none, and sharper questions where an answer worries you.

    Use it after the first management meeting, before you commit time to full diligence.

    What you enter

    • The deal: sector, stage and business model. Later stages treat missing data on retention, concentration, pricing, payback and margin as a red flag. The business model and regulated sectors add their own requests.
    • Thirteen checks: retention, usage depth, customer concentration, roadmap control, activation, pricing power, product leadership, engineering capacity, exposure to AI assistants, payback on acquisition, gross margin, security and compliance, and platform dependence.
    • Numbers, if you have them: gross and net revenue retention.
    • Each has three answers from good to concerning, plus a fourth for when you don't have the data yet.

    What you get

    • A status for each check: clear, watch, red flag or missing data. Each status has a word as well as a colour.
    • An overall rating of low, medium or high product risk.
    • Red flags in plain words, and why each matters.
    • An information request you can copy into an email.

    Scoring and assumptions

    Good answers score 2, middle answers 1 and concerning answers 0. Missing data scores 1, so a gap neither helps nor hurts, but it adds a request. Retention and AI exposure weigh double, and usage depth weighs one and a half. The weighted score sets the rating: 75% or more is low risk, 50 to 74% is medium, and under 50% is high. Steeply falling retention makes the rating high whatever the rest says. Five or more missing answers lead the request with a note that the picture is incomplete.

    The thresholds are rules of thumb for software companies. Adjust them to the sector and the stage. Usage benchmarks: Zylo's 2026 SaaS Management Index finds the average organisation uses 54% of the licences it pays for.

    See answer