On this page
Write a definition and reporting period for every metric. A small, auditable dashboard is more useful than impressive numbers nobody can reproduce.
Define the first useful customer outcome
A signup tells you that an account exists, not that the customer received value. Choose an activation event that represents the core job being completed. For a fictional reporting tool, that could be generating and reviewing the first usable report. Record the eligible account group, time window and whether a founder completed the task on the customer's behalf.
Illustration: if eight of twenty eligible new accounts complete that event within seven days, activation is 8 / 20, or 40%, for that cohort. This is a worked example, not a target. Compare subsequent cohorts with the same definition, and investigate the steps where accounts get stuck before celebrating a change in the percentage.
Separate recurring value from cash receipts
For a fixed-subscription model, monthly recurring revenue normalises recurring subscription amounts to one month. A EUR 1,200 annual subscription contributes EUR 100 to this simplified MRR calculation, not EUR 1,200. Exclude one-off implementation fees and taxes. Annualised recurring revenue in this simple model is MRR multiplied by twelve; it is not a forecast of cash collected.
Stripe documents its own billing definitions and configurable treatment of discounts and active subscribers. Align your definitions with your actual reporting tool, especially for trials, overdue subscriptions and usage-based revenue. Keep invoiced amounts, recognised revenue and cash receipts separate. Record exceptions rather than silently switching definitions when preparing an investor update.
Follow the same customer cohort over time
Retention asks whether the customers who started together continue to receive value or pay. Define whether you are measuring account activity, paying customers or recurring revenue. Use a cohort table to avoid confusing new acquisitions with the behaviour of existing customers. A growing total account count can hide a group that stops using the product after its first week.
In a simplified monthly example, two cancellations among twenty paying customers present at the start produce a customer churn rate of 2 / 20, or 10%. Exclude customers acquired during the month from that starting denominator. State how reactivations and scheduled cancellations are treated. With only twenty customers, one account changes the result substantially, so show counts alongside percentages.
Explain recurring revenue changes
Break movement into new subscriptions, expansion, contraction and lost recurring revenue. Net revenue retention looks at the starting customer group only: starting MRR plus expansion, minus contraction and churn, divided by starting MRR. Keep new-customer revenue out of this calculation and state how you handle reactivations, currencies and other exceptions.
Worked example: a cohort starts at EUR 2,000 MRR, adds EUR 300 through expansion, loses EUR 100 through downgrades and EUR 200 through cancellations. Net revenue retention is (2,000 + 300 - 100 - 200) / 2,000, or 100%. A separate EUR 500 from new customers grows total MRR but does not improve this cohort's retention. This example is not a benchmark.
Use acquisition costs with caution
Customer acquisition cost divides the relevant sales and marketing costs by the number of new customers acquired. Define the included costs and the attribution period. Long sales cycles create a timing problem: this month's spending may produce customers later. For founder-led sales, record time spent even if you also show a cash-only calculation, so that unpaid labour is not invisible.
Illustration: EUR 900 of specified acquisition spending and three new paying customers gives EUR 300 cash acquisition cost under that definition. It does not include every company cost or establish long-term profitability. Avoid confident lifetime-value estimates when you have only a few months of retention data. Use a scenario range and make its assumptions explicit.
Build a weekly and monthly review
Review onboarding failures and customer feedback weekly, then reconcile subscription movements and cash monthly. Give each metric an owner, source and calculation. Keep a short change log when the definition or product changes. Investigate missing events and duplicate accounts before interpreting a sudden jump as real business growth.
Finish the review with one action and a question it should answer. If activation is weak, improve a specific onboarding step. If customers activate but do not return, revisit the recurring job. If retention is strong but delivery costs are high, review scope and pricing. The dashboard should point back to customer behaviour, not become a substitute for talking to customers.
- Metric definition, source and reporting period.
- Raw counts and a consistent customer cohort.
- Changes since the previous review and known data limitations.
- One next action with an owner and review date.
Sources and review
Official sources checked on 6 October 2026. Programme terms and schedules can change. Check the provider's website before acting. Worked examples are illustrative, not client results.
