Renewal and retention in UK insurance broking: what the lapse rate really measures

Renewal, retention, attrition and tenure as the FCA defines them, what its home and motor data show, and how a UK broker should count its own book.

By FCB.ai, ORIS editorial teamEditorial lead: The ORIS editorial teamPublished on 9 min read
Contents
  1. The vocabulary: renewal, tenure, attrition, retention
  2. What the FCA's home and motor data show
  3. Why retention works differently since January 2022
  4. How to measure retention in your own book
  5. Messages that change the outcome before the renewal date
  6. Pitfalls in the way brokers read retention
  7. Where ORIS fits
  8. Frequently asked questions

In UK general insurance, a renewal is the carrying forward of a contract at expiry between the same parties, and retention is the share of expiring policies that are renewed; its mirror, the attrition or lapse rate, is the share of policies in force one year that are not renewed the next. On that measure, the FCA's own data put attrition at 25% for home and 40% for motor in 2022.

Those definitions sound obvious until a brokerage tries to compare its figure with an insurer's, or with last year's. Insurers, brokers and the regulator count different things under the same word, and since the pricing rules of January 2022 the levers that used to move retention are no longer available. This guide sets out the vocabulary as the FCA uses it, the market figures it has published, and a way for a broker to measure its own book that survives scrutiny.

The vocabulary: renewal, tenure, attrition, retention

The FCA's evaluation of its pricing remedies, EP25/2 (July 2025), reproduces the Handbook glossary terms and explains how it measured switching. Retention as a ratio is not a Handbook term: it is the complement of the attrition the FCA measured.

TermDefinition used by the FCAWhat it means for a broker's count
RenewalCarrying forward a contract, at the point of expiry and as a successive or separate operation of the same nature as the preceding contract, between the same contractual partiesA policy moved to a new insurer at expiry is not a renewal of the old contract, even though the client stays with you
TenureThe number of years a customer has held their policy, including any renewal of the policyCount it per client and per policy; the two diverge as clients add or drop covers
New business customerA prospective customer for a policy where the policy being taken out is not a renewalA client remarketed to a new insurer is new business for that insurer, retained business for you
AttritionA customer had a policy for a defined risk in one year for a given firm, and they do not in the next yearThe numerator of your lapse rate: policies in force at the start, not in force a year later
Retention rateNot defined in the Handbook; the complement of attritionPolicies renewed or replaced within your book, divided by policies due for renewal in the period

"Lapse" itself is used loosely. In UK general insurance it usually means a policy that ends at expiry without renewal or falls away for non-payment; in life and protection it means premiums stopped. Write down which one your firm means before anyone reports a number.

What the FCA's home and motor data show

For EP25/2, the FCA measured attrition as the percentage of policies in effect in 2020 not renewed in 2021 (before the remedies), and in effect in 2022 not renewed in 2023 (after). It treats attrition as a close proxy for switching, while acknowledging that a small proportion of consumers cancel and do not take out a new policy.

Indicator (FCA data, consumer home and motor)HomeMotor
Overall attrition, 2020 to 202226% to 25%39% to 40%
Attrition post-2020, policies without auto-renewal32.2% to 38.5%49% to 54%
Attrition post-2020, policies with auto-renewal19.6% to 21.5%33% to 35%
Share of policies held for 4 to 6 years, before and after the remedies15% to 18%11% to 14%
Policies with auto-renewal at inception, before and after54% to 70%74% to 77%

Three readings matter for a brokerage. First, these are insurer-level figures: a client who leaves one insurer for another through the same broker counts as attrition here and as retention in your book. Second, the gap between policies with and without auto-renewal is wide in both markets, which is why the cancellation and renewal mechanics deserve as much attention as the price. Third, the direction changed by tenure: attrition rose among lower-tenure consumers and fell among higher-tenure consumers, which the FCA reads as fairer prices removing the reason for loyal customers to shop around.

Why retention works differently since January 2022

The rules finalised in PS21/5 came into effect on 1 January 2022. ICOBS 6B.2.1R states that a firm must not set a renewal price that is higher than the equivalent new business price, and the FCA estimated that consumers in home and motor would save £4.2bn over a 10-year period. EP25/2 found the home gap narrowed: an existing policyholder paid on average £95.38 more than a new business customer before the reforms, and £49.17 more after.

Two points reach intermediaries directly. The rules apply to intermediaries involved in setting any portion of the renewal price, including where the intermediary rebates commission; the mere receipt of commission determined by another firm does not make a broker price-setting. And ICOBS 6A.6.4R requires the methods for cancelling auto-renewal to include at least all the methods by which a consumer can buy a new policy with the firm. A brokerage that sells over WhatsApp has to accept auto-renewal cancellations over WhatsApp.

The consequence for retention is that a home or motor client can no longer be kept by a discount that a new client would not get, or by making it awkward to leave. What remains is service: an accurate renewal, explained before the notice arrives, and a cancellation handled as quickly as a sale. For the disclosure side of the same moment, see our guide to what an ICOBS 6.5 renewal notice must contain.

How to measure retention in your own book

  1. Fix the cohort. Take every policy whose renewal date falls in a given month, frozen at the start of that month. Policies written later belong to the next cohort, so the denominator never moves.
  2. Count twice: by policy and by premium. Policy count tells you how many clients you kept; premium retention tells you what the book is worth. A book can keep its count while losing its largest accounts.
  3. Classify every policy not renewed. Client chose to leave; non-payment; risk ended (car sold, house sold); insurer declined to renew; broker declined. Only the first is a retention failure in the usual sense, and mixing them hides the cause.
  4. Treat remarketing as retained. Use the FCA definition: the old contract is not renewed, but the client stays. Report "renewed with the same insurer" and "moved within the book" as two lines of the same retained total.
  5. Split by tenure. In its policy statement the FCA noted that, for motor, the average tenure of consumers younger than 45 is less than 2 years and more than 4 years for those aged 65 and above. Your first and second renewals and your long-tenure clients behave differently; one average hides both.
  6. Split by auto-renewal. Given the gap in the FCA data, report retention separately for policies that renew automatically and those where the client must act.

Kept month by month, this gives a board report something better than a single percentage: where the book is leaking, and why. It is also the kind of outcome evidence the Consumer Duty expects a firm to monitor, discussed in our guide to fair value and broker commission evidence.

Messages that change the outcome before the renewal date

The policy that lapses quietly is usually the one where nobody spoke to the client between the last renewal and this one. Short, specific messages from the firm's number do more than a longer email. Examples for a personal-lines book:

Before the renewal is prepared: "Hello {first_name}, your home insurance comes up for renewal on {date}. Before we prepare it, has anything changed: an extension, new valuables, someone moving in or out? A one-line reply is enough. {broker_name}, {firm}"

After the renewal pack has been sent: "Hello {first_name}, your renewal documents for {policy} are in your inbox. If the price or the cover has changed and you want it explained, reply here and we will go through it with you."

When the client says they want to leave: "Understood, {first_name}. We have stopped the automatic renewal on {policy} and will confirm it in writing today. If it would help, we can also check the market for you before the cover ends."

After a policy has not renewed: "Hello {first_name}, we see your {policy} did not renew. If you are now insured elsewhere, nothing to do. If not, tell us and we will look at it before you are left without cover."

The third message is the one firms get wrong: the cancellation is actioned first, and the offer comes after, never as a condition. The full sequence is set out in the policy renewal use case.

Pitfalls in the way brokers read retention

  • Benchmarking against insurer figures. The FCA rates are per firm and per risk; your book retains clients the insurer lost. Compare yourself with your own previous cohorts.
  • Counting non-payment as a client decision. A failed instalment on premium finance is a servicing problem with its own fix, not a lost sale.
  • A loyalty discount at renewal. If your firm sets any portion of the home or motor price, a renewal price above what a new client would pay breaches ICOBS 6B, whatever it is called.
  • Slowing a cancellation to save it. A request sent on the channel the client bought through must be accepted on that channel.
  • One figure for the whole book. Without the split by tenure and by auto-renewal, a falling rate cannot be explained or fixed.

Where ORIS fits

ORIS works on the conversation side of retention. In Customers & Segments, the book can be filtered by renewal window, risk score and engagement score, so the pre-renewal check-in goes to the right cohort; lifecycle triggers fire a renewal reminder a set number of days before the date, from Meta-approved templates. Incoming replies are classified, and a message signalling a likely departure is flagged as high churn risk and appears in Opportunities & Risks for a handler to call. The cohort and reason counts above still come from your own systems: ORIS does not connect natively to a broker management system and exports customers, opportunities and campaign performance as CSV. You can see it on your own book via a demo.

Frequently asked questions

What is the difference between renewal rate and retention rate?

Renewal rate usually counts policies renewed with the same insurer; retention rate counts clients or premium kept in the book, including policies moved to a new insurer at expiry. Under the FCA's definition, moving insurer is not a renewal of the old contract, so a broker's retention rate should be at least as high as its renewal rate.

How does the FCA measure attrition?

In EP25/2, attrition is the percentage of policies in effect in one year that were not renewed in the following year with the same firm. The FCA used it as a proxy for switching, noting that a small proportion of consumers cancel and do not buy a new policy.

What retention rate should a UK broker aim for?

There is no regulatory benchmark, and the FCA's published figures are insurer-level rates for consumer home and motor, not broker retention. The useful target is your own trend by cohort, split by tenure and auto-renewal, with every lost policy given a reason.

Can a broker offer a renewal discount to keep a client?

For consumer home and motor, a broker that sets any portion of the renewal price cannot make it higher than the equivalent new business price, and the FCA's rules reach intermediaries who rebate commission. A broker can always remarket, adjust cover to the client's needs or explain the price; it cannot charge a loyal client more than a new one.

Do these rules apply to commercial renewals?

No. ICOBS 6B covers home and motor insurance sold to consumers. Commercial retention is still worth measuring the same way, by cohort and by reason, but the pricing rule does not apply to it.

See ORIS in action

Shared WhatsApp inbox, client records, follow-ups and opportunities for the whole brokerage. 15-minute demo.

Book a demo

Free kit : WhatsApp template pack for insurance brokers

This guide and its companions as a printable PDF, with a sourced checklist up front. 15 pages · PDF · 601 KB.

See the kit
Book a demo