Professional Services · Scenario 18

Insurance Agency

Most of the agency’s growth is sitting in its own book — clients with one policy who should have two. We model who is most likely to buy the second and stop spraying everyone.

Method · Uplift / propensity modelling

The situation

An insurance agency’s cheapest premium is a cross-sell to an existing, trusting client — yet most agencies blast the same ‘bundle and save’ message to the entire book, annoying the unlikely and underserving the ready. Renewals come and go without the one well-timed, relevant offer that would have landed.

Nobody scores which auto client is ripe for home, life or umbrella cover, so the agency’s richest growth channel — its own customers — is worked at random.

1.3
policies per client
Spray-all
the cross-sell method
At renewal
the missed moment
No score
of who will buy

Where we dig for the truth

We score every client for their propensity to take each additional product, and target only the moments and people most likely to convert.

Policy & coverage historyClient life-stage signalsRenewal timingClaims & contact historyProduct-holding patternsPremium & margin by product
Cross-sell conversion by propensity decileShare who buy the second policy when offered0%9%18%27%37%2%1-2 (low)5%3-49%5-617%7-831%9-10 (high)

The top two deciles convert at six-to-fifteen times the bottom. A propensity model tells the agency exactly whom to call, and when.

Our approach — Cross-Sell Propensity Modelling

An uplift / propensity model ranks clients by their likelihood to take each next product, scored on life-stage, current coverage and behaviour. The agency contacts high-propensity clients at the right trigger — a renewal, a life event — with the one relevant offer, and leaves the rest alone.

Producers’ time, the scarce resource, is pointed at the conversations most likely to bind a second policy, lifting policies-per-client without buying a single new lead.

From spray-all to scored cross-sell1Profile the bookAssemble coverage,life-stage andbehaviour for everyclient.2Score propensityModel who is mostlikely to buy each nextproduct.3Time the offerTrigger the right offerat renewals and lifeevents.4Focus producersAim agent time at thehighest-propensityconversations.
Cross-sell funnel after scoringTargeted clients to bound policiesSTEP RATEHigh-propensity clients600Contacted with relevant offer54090%Quoted24345%Second policy bound13857%

Targeting only the ready clients turns a low-yield blast into a funnel where nearly a quarter of those quoted bind a second policy.

What changes

Same book, worked by propensity. Representative for an independent multi-line agency.

Representative 90-day movementPolicies / client1.31.7▲ +0.4Cross-sell close6%23%▲ +17 ptsClient retention84%91%▲ +7 ptsAnnual revenue$1.4M$1.7M▲ +21%
Where the growth comes from+$0.3Mannual revenueTargeted cross-sell52%Higher retention (bundles)30%Better-timed offers18%
Why this is not "social media management"
We didn't buy the agency more cold leads. We mined its own book for the second policy already waiting in it, and aimed producers at the clients most likely to say yes. Propensity modelling grows revenue from customers you already have.

Frequently asked questions

How do you increase policies per client?
We score every client for their propensity to take each additional product, then contact only the high-propensity clients at the right trigger — a renewal or life event — with one relevant offer, instead of blasting the whole book.
What is propensity modelling?
A propensity model uses a client's profile and behaviour to estimate how likely they are to take a specific action, such as adding home cover to an auto policy. It points producers at the conversations most likely to close.
Isn't buying leads faster?
Your cheapest premium is a cross-sell to an existing, trusting client. We mine your own book for the second policy already waiting in it. Book a marketing audit.

Want this run on your numbers?

Share your policy book and we’ll score where the second policies are hiding.