Why the numbers carry so much weight
A subscription business is valued on its metrics more than on its accounts, because the metrics describe how the revenue behaves and the accounts only describe what it was last year.
That gives sellers an unusual amount of control over how they are perceived, and an unusual amount of exposure. Metrics presented inconsistently, or defined generously, are the fastest way to lose a buyer's confidence in everything else, because these are the numbers they will rebuild themselves from raw data.
The revenue metrics
Annual recurring revenue is the base of most valuations. What matters is the definition. Buyers want to know whether it includes implementation fees, professional services, usage overages, or contracts within their notice period. Every one of those has been included by someone at some point, and buyers now assume the broadest definition until shown otherwise.
Monthly recurring revenue matters mainly for showing trend and seasonality, and for spotting the month the growth actually changed.
Net revenue retention measures revenue from an existing cohort against the same cohort a year earlier, including expansion, contraction and churn. Above 110% the business grows without adding a customer. Above 130% is rare. This is the metric with the strongest relationship to the multiple achieved.
Gross revenue retention takes expansion out and shows pure leakage. Above 90% is good and above 95% is strong. The gap between net and gross retention is one of the most revealing numbers in a diligence pack, because a healthy net figure can conceal serious churn masked by expansion in a handful of large accounts.
Logo churn counts customers rather than revenue. A business can hold net retention above 110% while losing a third of its small customers, and that pattern says something specific about product fit in the lower segment which a buyer will want explained.
The unit economics
Customer acquisition cost, fully loaded. Sales and marketing salaries, commission, advertising, tooling and the share of overhead that supports them. Calculated by channel and segment rather than in aggregate, because the aggregate hides the channel that stopped working.
Lifetime value, being average revenue per account divided by churn rate and multiplied by gross margin. The gross margin step is often omitted, which inflates the figure and is noticed.
The ratio between them. Above three suggests each customer returns considerably more than they cost. Below one is unsustainable and visible immediately.
Payback period, meaning months to recover acquisition cost. Under twelve is strong for a middle-market business, under eighteen acceptable. This has become more important as capital has become more expensive, because a long payback means growth that has to be funded for longer before it returns anything.
Cohorts, which is where buyers actually look
Aggregate metrics can be flattered by mix. Cohort data cannot.
Revenue cohorts show whether customers acquired in a given period spend more or less as they age. A cohort curve that flattens after year two tells a different story from one that keeps climbing.
Retention cohorts show whether churn is improving. A business with weak overall retention but visibly improving cohorts is a different proposition from one where each cohort is worse than the last, and the aggregate number does not distinguish them.
Segment cohorts show where the value concentrates. Frequently the enterprise segment is excellent and the small business segment is loss-making, and the aggregate hides both.
Expect a serious buyer to rebuild all three from the raw export rather than accept a summary.
Presenting them
One definition per metric, used identically in the memorandum, the data room and the management presentation. A metric that changes definition between documents is treated as evidence that the favourable version was chosen each time.
Show the raw data underneath. A buyer who can verify the arithmetic stops questioning it.
Benchmark honestly against public medians rather than against best in class. Comparing a middle-market business to the strongest listed companies invites the obvious response.
Show trend even where the absolute level is unimpressive. Retention improving from 95% to 104% over six quarters is a stronger story than a static 108%, and it is a story the cohort data supports.
Segment everything. Buyers assign different values to different parts of a business, and a seller who has done that work controls the framing rather than reacting to the buyer's version of it.
The one that decides the price
If only one number were available, it would be net revenue retention, with gross retention beside it for context. Everything else in this article describes the business. Those two describe whether it keeps what it has, and that is what a buyer is ultimately underwriting.
Editorial Team · Published for orientation, not as advice on a specific transaction. Any figure cited is orientation, not a valuation. See market notes.