In brief
- ARR measures scale, not quality - two businesses with identical ARR can be worlds apart.
- Net revenue retention, cohort behaviour and unit economics reveal the real durability.
- The question is not how fast a business grew, but how much of that growth it keeps.
The metric that hides the most
Every software business leads with annual recurring revenue, and for good reason - it is a clean measure of scale. But ARR describes how big a business is, not how good it is. Two companies with identical ARR can have entirely different prospects depending on how that revenue behaves once it is won. In diligence, ARR is where the conversation starts, not where it ends.
Retention is the real story
The single most revealing signal is net revenue retention - what a cohort of customers is worth a year later, after churn, contraction and expansion. A business that retains and expands its existing base compounds; one that must replace lost revenue just to stand still is running to keep up, however fast headline growth looks. Gross and net retention, read together, tell you whether growth is durable or merely purchased.
The question is not how fast a software business grew. It is how much of that growth it gets to keep.
Cohorts do not lie
Aggregate numbers flatter; cohorts expose. Following successive vintages of customers over time reveals whether the product is getting stickier or weaker, whether newer customers behave like the profitable early adopters or are being acquired on thinner terms, and whether expansion is broad-based or concentrated in a handful of accounts. Concentration - in customers, in a single expansion motion, in one channel - is a risk that ARR conceals entirely.
Unit economics and the cost of growth
Durable software businesses earn back what they spend to acquire customers, and do so within a sensible period. The relationship between acquisition cost and customer lifetime value, the payback period, and the efficiency of each incremental sales pound separate businesses that can grow profitably from those buying growth at any price. A high growth rate funded by uneconomic acquisition is a liability wearing the costume of an asset.
Reading the whole picture
No single metric decides quality. Retention without healthy unit economics is fragile; strong unit economics without retention is a leaky bucket. The judgement lies in reading them together - and in testing the story against the granular data rather than the pitch. That is where a genuinely durable software business reveals itself.
How Queen's Tower helps
Our TMT and data specialists interrogate the metrics that matter - cohort behaviour, retention, concentration and unit economics - to separate durable software businesses from those that simply look good on an ARR chart.
