1 · Credibility blend
Slide between own experience and the benchmark
live model
The credibility estimate blends the observed rate with the expected benchmark. More exposure raises Z, so the model puts more weight on the company's own experience.
Expected claims
1.9
benchmark x exposure
Credibility Z
22%
weight on own data
Estimate
4.6 /100
credibility-weighted rate
Credibility breakdown
Synthetic establishment · Selected benchmark · 60 employees
benchmark 3.2
own data 9.8
estimate 4.6
rate per 100 FTE — the estimate is pulled toward whichever side has the data
Own rate
9.8/100
Benchmark
3.2/100
Cred. estimate
4.6/100
Credibility Z — weight on own data22%
trust the benchmarkE = 1.9 claimstrust the company
Only 22% credibility — too thin to trust on its own; the estimate barely leaves the benchmark. It reads Insufficient despite the raw rate.