Home / Commercial / Pilot to scale
A first step that answers a question, not a first step that buys time.
A pilot is worth running only if it is designed to close a specific gap in the evidence. Ours is built around the two assumed links in the loss ratio chain and the unmeasured conversion funnel.
Scope
A defined household cohort in a single market, one entry line (motor), one panel, one deployment route, with a control group where the partner can support one.
Duration
Long enough to observe renewal events and behaviour change, which in practice means two to three renewal cycles of observation rather than a ninety-day sprint. The engagement and conversion metrics report far earlier.
Investment
Capped, with the split between partner and Amanati set in the pilot agreement. Structures vary by route — in the agent route the insurer funds deployment; in the broker route the broker does.
Success criteria
Agreed before launch, in writing, with the measurement method specified. Vault completion, T-30 engagement, quote and conversion rates, renewal events controlled, and where the cohort size permits, an early claims signal.
Decision gate
A single go or no-go against the criteria, with the data handed over in full either way. A pilot that fails its criteria should end.
What the pilot is designed to prove
That households will upload documents at a rate that makes the vault useful. That the T-30 trigger produces engagement rather than annoyance. That reason-coded recommendations convert. And, over a longer horizon, that scored feedback changes driving behaviour enough to move claims.
The first three report within months. The fourth is the one a reinsurer cares about and it takes longer — which is why the commercial case should not wait on it.
Phase 0 blockers
Three things are resolved before engineering begins on any deployment: out-of-city assistance fulfilment, regulatory clearance for the market, and telematics approach.
Starting a build with any of these open produces a pilot that stalls for reasons that have nothing to do with the product.
Scope a pilot for your market.
Bring the constraint you are most worried about. That is usually where the design should start.