6 min read · 927 words
At a glance
- The ratio mirage is what happens when a growth-stage company’s headline ratios stay inside acceptable bands but the denominators quietly drift. The number holds. The signal is gone.
- Three drivers: customer definition creep, cohort mixing, and pricing-tier consolidation. Each one is invisible inside the monthly report.
- The 14-day fix is not better dashboards. It is restating the three ratios the board actually watches, cohort by cohort, and forcing one explicit definition per term.
The first sign a post-Series A or Series B company has stopped tracking reality is not a bad number. It is a reassuring one. Gross margin at 78%. LTV/CAC at 4.2x. Net revenue retention at 108%. Every ratio inside the band a board expects to see. Nothing to flag. Nothing to discuss. Nothing to do.
That is the mirage. The ratio looks healthy because the formula still runs. The formula still runs because the denominator has quietly drifted out from underneath it. The number is true, arithmetically. It is false as a signal.
How does the ratio mirage form?
Three drivers, always in combination.
Customer definition creep
Between Series A and Series B, the definition of “customer” almost always changes. A free tier is added. A self-serve path launches. A partner reseller becomes a meaningful share of logos. An enterprise motion starts closing multi-team deals that used to be one team.
Every one of those changes is correct on its own. The problem is that nobody updates the retention calculation, the CAC calculation, or the LTV calculation when the definition moves. The ratio keeps computing against a customer population that no longer means what it did eighteen months ago. Net revenue retention of 108% against a mixed population of free-tier upgrades, paid cohorts, and partner-resold seats is not the same number as 108% against paid new-logo enterprise. One is a real signal. The other is a comfortable average.
Cohort mixing
Cohorts are supposed to be the fix for this. In a walking dead portco they almost never are. The monthly board report shows a blended retention number. The cohort detail lives in a tab nobody pulls up. The tab was built two CFOs ago. The formulas reference sheets that have since been renamed. Nobody trusts it enough to cite it, so it does not get cited, so it does not get fixed.
What you get instead is a single-line ratio that averages a healthy early cohort against a struggling current cohort. The healthy early cohort is a survivor bias artifact; the current cohort is what predicts the next year. You cannot see the current cohort in the average. You can only see the average.
Pricing-tier consolidation
At some point the company rationalizes pricing. Usually around month six post-Series A and again around month three post-Series B. Old tiers get grandfathered, new tiers get launched, some customers get moved onto multi-year contracts to lock them in. The gross margin calculation does not always get restated for the shift. If 20% of revenue is grandfathered at a lower price point with full product access, and another 15% is on a lower-margin pricing plan that was introduced to rescue a segment, the blended gross margin is a fiction.
The CFO knows this. The CFO has not had the bandwidth to restate it. Nobody has asked because the blended number is still inside the band.
What do you do about it?
The fix is not a better dashboard. It is three deliberate acts of plumbing, done once, inside two weeks.
- Restate the three headline ratios against one explicit customer definition. Pick the definition you actually want to track, whether that is paid new-logo, paid active, or paid active of target size. Recalculate gross margin, LTV/CAC, and NRR against only that set. The new numbers will almost certainly look worse than the blended version. That is the point.
- Break out the current cohort from the historical average. Anything older than 18 months goes in its own line. Anything newer goes in another. If the current cohort’s retention is materially below the historical average, that gap is the conversation.
- Footnote every ratio with its denominator. Every headline ratio on the board deck gets a one-line denominator definition directly below it. “NRR of 108% against paid new-logo accounts with $50K+ ACV, excluding grandfathered pricing.” If the definition is too complicated to fit on one line, the ratio is not ready to report.
Why this is worth doing the week you read it
The ratio mirage is not a reporting problem. It is a decision problem. When the ratios look fine, nothing triggers a strategic review. The company keeps spending against a plan that was built on a denominator that no longer exists. Burn rises. Revenue creeps. Six months later the raise stalls and everyone is surprised.
The companies we run the Diagnostic on are never surprised by their own numbers after the first 14 days. Most of them are surprised by how different the restated numbers look. That is the first, hardest step out of the walking dead pattern: being willing to see the numbers as they actually are.
Where to go from here
If any of this tracks, the six-signal pattern is probably running somewhere in your company. The self-assessment takes six minutes and names which signals are live.
Related reading: