An agency calls every renewal above $25,000 in premium and sends everyone else a letter. Aimed by the chance of losing the account times its commission, the same calls cover $243,800 of the commission that left, against $140,400.
The agency principal decides which renewals get a call from an account manager and which get the letter 60 days out. The rule is to call every account above $25,000 in premium. It feels fine because the big accounts are where the commission is. The rule never asks which accounts are likely to leave.
We name the baseline. The $25,000 premium line stays in every comparison, at the same 480 calls a year the account managers can make.
We write down the expectation before the data is opened. Rate increases and claims should carry most of the risk, and the score should beat the premium line on commission that leaves.
We build on two years of renewals with their outcomes. Each account gets the chance it does not renew, times its commission.
We freeze the score and write down a forecast of $231,000 of lost commission on its call list. Then we run it once on the next year's 2,400 renewals. The list holds $243,800, 5.5% over the forecast.
We try the simpler ways. A sort on rate increase and a sort on last year's claims are each scored at the same 480 calls. The score is 23% to 74% ahead of them and of the premium line.
We hand back the score as a field in the agency management system, the monthly call list it sorts, the memo, and the holdout design.
Of its 480 calls, 270 go to accounts the score rates low risk, and only 2.2% of those leave.
So what More than half the call capacity goes to renewals that were rarely in doubt.
Accounts renewing with premium up more than 15% leave at 17.8%, against 7.4% for the rest.
So what The renewal itself tells the team who to call, months before the letter goes out.
It holds 52.6% of the $463,400 lost on the judged year, against 30.3% for the premium list.
So what The same people making the same number of calls reach the accounts where a call has something to save.
The account managers work a monthly call list sorted by the score, 120 days before renewal, at the same capacity as before. One in eight accounts on the scored list, about 60 a year, gets the letter only, chosen at random. Comparing their renewal rate with the called accounts measures what a call actually saves, which no amount of history can.
The scored list holds more of the commission that left. Whether a call keeps it is unknown until the holdout cell reads out.
Large accounts were always called, so some of their risk was already worked down. The score may rank them lower than it should, which leans against the premium line in the comparison.
Sixty accounts a year is a small cell. The first year gives a direction, not a verdict.
We'll tell you what your own history can say about it, and which offering fits. If none does, we'll say that too.
Start a conversation