Anonymizing Insurance Claims Data
What it takes to anonymize claims data, how that differs from pseudonymization and masking, and why removing names is only the start of the job.
Glossary
Insurance Retention Rate
How insurers, agencies and brokers measure retention, the difference between policy, customer and premium retention, and how to calculate each one.
Insurance retention rate measures how much existing business stays with an insurer, agency, or broker over a defined period. It can track renewed policies, retained customers, or the premium associated with continuing business.
These measures answer different questions. A company can retain most of its customers while losing policies, or increase renewal premium while retaining fewer customers. A useful retention figure makes clear what is being counted and which group is being measured.
What is retention rate in insurance?
Retention describes the continuation of an existing insurance relationship. In an annual motor policy, that might mean the policyholder renews for another year. For an agency serving businesses, it might mean the client keeps at least one policy with the agency.
The meaning also depends on whose business is being measured. If a broker moves a customer’s cover from one insurer to another, the broker may retain the customer while the original insurer loses the policy.
Published definitions vary. Travelers, for example, describes commercial insurance retention using premium available for renewal, excluding rate and exposure changes, while its personal insurance definition uses renewal policy counts. Travelers’ retention definitions.
Here, retention refers to keeping existing business. It is separate from an insurer’s risk retention, which describes the risk it keeps rather than transfers through reinsurance.
How do you calculate policy retention rate?
A straightforward renewal-cohort calculation is:
Policy retention rate = policies renewed from the measured group ÷ policies due for renewal in that group × 100
The group, sometimes called a cohort, consists of policies with renewal dates in the chosen period. The numerator must come from that same group.
Policy retention example
Suppose a fictional insurer has 1,000 annual policies due for renewal in June. Of those policies, 880 renew.
880 ÷ 1,000 × 100 = 88% policy retention
The insurer also writes 200 policies for new customers during June. Those policies do not enter this retention calculation. They contribute to growth, but they do not show whether the existing book stayed.
The distinction explains why a growing policy count can conceal a retention problem. New sales may replace departing customers quickly enough to increase the overall book.
Which policies belong in the denominator?
The measurement needs a consistent rule for policies that the insurer decides not to renew.
If all 1,000 expiring policies are counted, the result measures retention of the expiring book. If 50 receive no renewal offer and the denominator includes only the remaining 950, the same 880 renewals produce a 92.6% acceptance rate among offered renewals.
Both figures can be useful, but they describe different outcomes. Labeling the second figure as renewal-offer acceptance makes the difference visible.
Public reporting illustrates this variation: Neptune defines its policy retention rate around policyholders who receive and accept renewal offers. Neptune’s operating metric definitions.
Policy retention vs customer retention
Policy retention counts contracts. Customer retention counts customer relationships.
A customer with home and motor insurance may renew the home policy and move the motor policy elsewhere. For that customer, policy retention is 50%, while customer retention is 100% under a definition that requires at least one policy to remain active.
A common customer-cohort formula is:
Customer retention rate = starting customers still active at period end ÷ customers active at period start × 100
New customers are excluded from the numerator. The company also needs a stable customer identifier so that a changed email address or replacement policy number does not create a false departure.
Goosehead’s published client-retention definition follows the relationship approach: it compares clients who had at least one policy in force a year earlier with those who still have at least one policy at the measurement date. Goosehead’s client-retention definition.
What is premium retention?
Premium retention measures the value of continuing business. The formula depends on whether the insurer wants to isolate retention or include changes in price and exposure.
Premium retention at expiring values = expiring premium of renewed policies ÷ expiring premium of the renewal group × 100
This weights each policy by its original premium. Losing one large commercial account can therefore have a much greater effect than losing several small accounts.
The Hartford’s commercial premium-retention definition excludes renewal price increases and exposure changes, allowing the measure to focus on the premium base successfully renewed. The Hartford’s premium-retention definition.
How a price increase changes the picture
Return to the fictional group of 1,000 policies. Assume each originally had a $1,000 annual premium, giving an expiring premium total of $1 million.
Of these, 880 renew at a new premium of $1,100 each.
| Measure | Calculation | Result |
|---|---|---|
| Policy retention | 880 ÷ 1,000 | 88% |
| Premium retention at expiring values | $880,000 ÷ $1,000,000 | 88% |
| Renewal premium relative to expiring premium, including the price increase | $968,000 ÷ $1,000,000 | 96.8% |
The insurer has retained 88% of the policies. The higher 96.8% figure also captures the 10% premium increase. Presenting both makes it possible to distinguish customer continuity from pricing effects.
What counts as a lapse or lost policy?
A retention report needs clear treatment of policies that stop, change, or temporarily leave the book.
Common situations include:
- Customer non-renewal: the customer declines the next policy term or moves to another provider.
- Insurer non-renewal: the insurer decides not to offer a further term.
- Cancellation during the term: coverage ends before its scheduled expiry.
- Lapse following nonpayment: coverage ends under the policy’s payment and notice provisions.
- Reinstatement: a previously ended policy becomes active again under the applicable terms.
- Replacement or rewrite: a policy continues under a new reference or product structure.
A midterm cancellation affects an in-force customer measure immediately, but its treatment in a renewal-cohort report depends on the defined methodology. Reinstatements may also change a result after the initial reporting date.
For example, a rewrite should be linked to its original policy where it represents continuing business. Otherwise, the system may report one lost policy and one new sale even though the customer stayed.
What is a good insurance retention rate?
A meaningful target depends on the product, distribution channel, customer group, and measurement method. A renewal-offer acceptance rate cannot be compared directly with a measure that includes insurer-initiated non-renewals.
Retention also needs to be considered alongside pricing and claims performance. Keeping business at an inadequate premium may raise retention while weakening underwriting results.
For operational use, the most helpful view separates retention by renewal month, product, channel, tenure, and reason for leaving. That makes it possible to distinguish a pricing issue from a service problem or a deliberate change in underwriting appetite.
A policy administration system can provide the renewal dates, transaction history, and links between policy terms needed for this analysis. Those records let teams trace the percentage back to the individual policies behind it.
