First, choose the book you are measuring
A growing premium total can coexist with a shrinking customer base. If prices rise while accounts leave, one headline number can hide two different stories. Begin with a defined cohort and a defined period rather than pulling unrelated dashboard totals into the same ratio.
For the examples below, the cohort is the set of customer accounts active at the start of the year. Track that same group at year end. Report new accounts separately. Decide how you handle acquired books, merged customer records, rewritten policies, seasonal business, and accounts that cease operations. Apply the definitions consistently and show exclusions.
The formulas here are internal management definitions. They are not insurer statutory accounting measures, audited financial statements, or industry benchmarks. Reconcile revenue and expense classifications with your finance team before using them for compensation or financial decisions.
1. Account retention: did the relationship stay?
Account retention = starting-cohort accounts still active at the measurement date ÷ starting-cohort accounts. Define “active” in advance—for example, at least one in-force policy serviced by the agency. Do not add new accounts to the numerator.
Worked example: 1,000 accounts are active at the start, and 900 of those remain active at year end. Account retention is 900 ÷ 1,000 = 90%. If 150 new accounts arrive, the ending total is 1,050, but retention is still 90%. Growth did not erase the 100 lost relationships.
Review departure reasons and the effort required to win replacement accounts. A loss due to a business closing calls for a different response from a loss following an unanswered service request. Keep both visible instead of removing inconvenient losses from the calculation.
2. Policy retention: what stayed within the relationship?
Account retention alone can miss a household or business moving most of its policies elsewhere while leaving one small policy behind. Use a second measure: policy relationships continued from the starting cohort ÷ policy relationships in that cohort.
Decide how to recognize a rewrite before counting. If a policy moves to another carrier through your agency, a canceled policy number may represent a continuing relationship rather than a loss. Maintain a documented predecessor/successor mapping so duplicate counting does not depend on memory.
Example: 1,800 starting policy relationships, with 1,620 continuing under the agreed definition, gives 90% policy retention. Investigate differences between the account and policy views rather than choosing whichever percentage looks better.
3. Cohort revenue: did higher rates disguise attrition?
Compare agency revenue from the same customer cohort across equal periods. Use commission and fee revenue under a consistent accounting basis, not written premium as a substitute for agency income. Identify changes in commission terms, exposure, policies purchased, and price where reliable data allows.
Example: the starting cohort produced $200,000 of agency revenue in the prior period. The surviving accounts produce $210,000 in the current period. The cohort revenue ratio is 105%, even though account retention is 90%. Both numbers can be true. The example alone does not prove whether rate changes, exposures, commissions, or additional business caused the revenue increase.
Show new-business revenue and unusual items separately. A one-time payment should not silently become evidence that the recurring book has improved.
4. Service effort: what did the revenue require?
For a small sample, record time spent on intake, research, customer contact, rework, and follow-up. Multiply recorded hours by an agreed loaded hourly cost. Keep the assumptions visible: salary alone is not the same as a fully loaded cost.
Illustration: an account generates $900 in annual agency revenue. Twelve recorded service hours at $45 per hour produce a $540 service-cost estimate, leaving $360 before acquisition costs and other overhead. That is a contribution estimate, not net profit. If hours are missing or estimated, label the result accordingly.
Use the findings to locate repeated work, poor intake, or preventable handoffs. Do not use a rough cost estimate as a reason to neglect contractual service obligations or treat clients unfairly. Improve the operating process before assuming the customer is the problem.
5. Follow-up reliability: did you keep the promise?
Follow-up reliability = commitments completed by their promised time ÷ commitments due in the period. Count an unresolved external dependency as unresolved; a reminder sent to yourself is not the promised customer update.
Example: 80 commitments were due this week, and 68 were completed on time. Reliability is 85%. Separate customer updates, document requests, licensed reviews, and external confirmations so the team can see which handoff is breaking down.
Do not silently move a deadline after it is missed. Preserve the original promise, record the reason, and make a new commitment the team can meet. Otherwise the metric rewards editing dates instead of doing the work.
Use a 20-minute weekly review
Start with overdue promises because those affect a customer now. Then review lost or partially lost relationships, recurring service bottlenecks, and the next renewal segment. Account and cohort revenue trends may be more useful monthly or quarterly; do not force noisy weekly movements into a narrative.
End with one named owner, one process change, and one date to inspect the result. For example: add a missing-information checklist to certificate intake, then compare the number of back-and-forth contacts on the next reviewed sample. A scorecard earns its place when it changes a decision.
- Write the cohort and measurement dates.
- Keep new business separate from retained business.
- Document rewrite and exclusion rules.
- Reconcile revenue definitions with finance.
- Show estimates and missing data.
- Assign one improvement owner and review date.
About the author
Abraham Cannon · Founder of PANDA OS
Abraham Cannon is the founder of PANDA OS. His focus is helping insurance agencies turn everyday sales, service, and renewal work into connected, accountable operations.