Churn Analysis

Churn analysis is the practice of examining which customers left or cut their spend, when it happened, and why, so you can find the specific pattern driving revenue loss and act on it. A single blended churn rate tells you almost nothing. The number that matters shows up only after you break churn apart by cohort, by reason, and by value.

Segment churn by cohort

Group customers by the month they signed up, then track how many stay active in each month afterward. This is cohort retention, and it separates two problems that a blended rate hides. If older cohorts retain well but recent ones drop fast, the product is fine and something in recent acquisition, onboarding, or pricing broke. If every cohort decays at the same point in its lifecycle, such as month 4, you have a structural retention problem tied to the customer journey rather than to one weak quarter of sales.

Reading retention this way also tells you where to intervene. A cliff at month 3 points to onboarding or first-value delivery. A slow bleed across every month points to ongoing value and price fit.

Segment churn by reason

Cohorts show you when. Reason shows you why. Tag every churned account with a cause: competitive loss, budget cut, missing feature, weak onboarding, champion departure, or price. Then rank those causes by the revenue attached to each, not by how many accounts fall under them.

Engagement data predicts churn earlier than any exit survey. ORM's customer data shows that support-ticket volume is one of the clearest signals. Accounts with zero support cases in a year are at risk, because silence usually means disengagement rather than satisfaction. Accounts with seven or more cases in a year are also at risk. The healthiest range is three to five tickets a year, usually tier 2 or 3 issues, which means the customer is engaged and getting help.

Segment churn by value

Counting logos treats a $2,000 account and a $200,000 account as equal losses. They are not. Weight churn by ARR and the priority list often flips.

This is where gross and net retention separate. Net revenue retention includes expansion, so a few large upsells can mask heavy churn underneath. Gross revenue retention removes expansion and shows only what you lost to churn and contraction. ORM builds a monthly reconciling waterfall from beginning ARR to ending ARR, splitting churned-customer ARR, churned-product ARR, and product downgrades on the loss side from new-customer and expansion ARR on the growth side. That breakdown tells you whether you are losing whole customers, losing products inside accounts you keep, or absorbing price downgrades.

Find the pattern worth acting on

Run all three cuts together and the real problem usually names itself. You find one cohort that leaves early, you can see the reason it leaves, and you can measure how much revenue that group represents. If the number is large enough, it earns a roadmap change or a dedicated retention play. Everything else is noise you should not staff against.