North Metric

    How we estimateprivate revenue.

    Private companies don’t publish revenue. We build a repeatable, testable estimate — and show the error bars.

    All estimates →

    Two independent views, reconciled

    There is no single public number that gives away a private company’s revenue. But there are two independent ways to see it — and when both agree, the range gets narrow.

    METHOD 01The website funnelReal acquisition traffic, run through published SaaS conversion rates and the company’s listed pricing. Best for self-serve revenue.
    Visits→Signups→Paid→× Price= Self-serve ARR
    METHOD 02The demand modelBrand search, company age and funding stage, fitted on companies whose revenue is public. Sees the sales-led revenue a funnel can’t.
    Brand search→Age→Stage= Total ARR

    We reconcile the two. For a self-serve product the funnel leads and the demand model confirms; for an enterprise company the demand model leads and the funnel is a floor. The result is one range.

    We look things up — we don’t invent them

    Every rate in the funnel is a published industry benchmark, not a knob tuned to get a nice answer. Only the demand model is fitted, and only on real reported revenue.

    RATEBENCHMARKSOURCE
    Visitor → signup0.4–2%Published SaaS funnel benchmarks, by business model
    Signup → paid3–30%Product-led growth and trial conversion reports
    Monthly logo churn1.4–4.5%Public SaaS retention studies, by stage
    Plan mix55/30/12/3%Weighted across the company’s own listed tiers

    Where a real number exists, we use it

    Some revenue has been credibly reported — press, filings or reputable data providers. Those pages show the reported figure, not our estimate. We never call an estimate “reported”.

    ✓REPORTEDReportedA figure from filings, press or a data provider. Shown as-is.
    HIGH CONFIDENCEBlendedA third-party estimate combined with our two methods.
    MEDIUM CONFIDENCEModelledOur two methods only, clearly labelled.

    How accurate is it?

    Not precisely. Every figure is modelled from public signals and published benchmarks, not from the company’s own books, so any estimate can be far off the real number. Each page shows a range and a confidence level for its figures, and a reported figure is marked as reported.

    It is tightest for self-serve companies, whose revenue shows up in web traffic, and loosest for enterprise companies, whose revenue is nearly invisible to public signals.

    Estimates are a range. Your Stripe is exact.Get your free churn report: money at risk, money leaking and the first 5 customers to act on.
    ACTIONS OFF BY DEFAULTGet your free churn report →

    Ranges and confidence, not one number

    Every estimate is a range with a confidence label. Read the range, not the midpoint.

    HIGH CONFIDENCETwo methods agree closely, or a reported anchor exists. Narrow range.
    MEDIUM CONFIDENCESolid public signal, or a third-party estimate blended in.
    LOW CONFIDENCEThin public signal. The range is wide on purpose — treat it as order of magnitude.

    What this is — and isn’t

    IT IS✓A transparent, testable estimate from public signals and published benchmarks.✓Updated as traffic, pricing and reported figures change.
    IT ISN’T✕Audited financials, insider information or a company-confirmed number.✕A precise figure — anyone quoting one exact number is guessing.
    Spot something off?Work at one of these companies? Send a correction and we’ll update the page.Request a correction →
    EXACT, NOT ESTIMATED

    Estimates are a range.Your Stripe is exact.

    Get your free churn report →Actions off until you turn them on