Net revenue retention measures what a group of existing customers is worth after a year, including expansion, downgrades and cancellations, with no new customers counted. It is expressed as a percentage of where that group started, so a figure above 100% means the existing base grew on its own.
Above 100% is the number every subscription business wants, and the reason is arithmetic rather than fashion. A base that grows without new logos means sales effort compounds: this year's new customers become next year's larger customers, and the company can grow while acquiring at a constant rate. Below 100% the opposite holds, and every month's new business is partly spent refilling a bucket that leaks. The measure deliberately excludes new customers so that acquisition cannot hide a retention problem, which is exactly what a headline revenue growth figure does.
Two readings of the same year are worth having side by side. Net retention nets expansion against contraction and churn, so a company can post 105% while losing customers steadily, provided the survivors spend enough more to cover the losses. Gross retention counts only the losses and can never exceed 100%, which makes it the honest measure of whether customers stay. A pair like 105% net and 86% gross describes a business with a churn problem funded by an upsell motion, and that is a materially different company from one at 105% net and 98% gross, even though the headline is identical. For a commission plan the distinction decides who gets paid: expansion is usually somebody's quota, renewal is usually somebody else's, and a plan that pays against net retention alone rewards the first team for the second team's work.
Take every customer holding a subscription on 1 January and measure only those customers a year later. New customers acquired during the year are excluded entirely.
Gross retention on this same cohort is $3,440,000 over $4,000,000, or 86%, because it ignores the expansion. One year, two measures, and a $760,000 upsell motion covering $560,000 of losses.
The measure is only as trustworthy as the rule for who is in the cohort. Drop a customer who churned in month two because they were never properly onboarded, treat a renegotiated contract as a new logo rather than as a downgrade, or count a customer that consolidated two subsidiaries as expansion when one subsidiary was already paying, and the figure moves several points without anybody writing anything untrue. Fix the cohort at the start date, follow every account in it to the end date including the ones that left, and write down how acquisitions, renamed entities and contract restructures are handled before the number is needed rather than after.
Commish pays net revenue retention the way your plan describes it, shows the arithmetic on every line, and traces each payment back to the deal that earned it.