Three live tiles from the same shop-level dataset behind the dashboard — total recovered calls, an after-hours vs business-hours split, and a per-trade breakdown. The numbers below answer from the exact same query paths the founder reads against, just without the caller phone or transcript. Refresh the page — they move.
Live · not a deck screenshot
Three live tiles from the same shop-level dataset behind the dashboard: total recovered calls, an after-hours vs business-hours split, and a per-trade breakdown. The numbers refresh on every page load.
Methodology
A recovered call is a call the AI voice agent answered and on which it confirmed homeowner intent for a real service need. These are live counts from each shop's call logs—not industry averages, modeled projections, or vendor reports. The after-hours versus business-hours split is keyed to the timestamp when the call was answered, not a first-touch webhook. The per-trade tile is an unfiltered count of jobs carrying a trade tag, not a back-out projection.
Also see
Each is short, on the same voice as the tiles above, and lands one click from the audit CTA.
Real transcripts
The six steps
Lost-job estimate
Next step
A 20-minute audit uses ten of your recent inbound calls and your last quarter's ticket values to fill the same three tiles for your shop — no install, no commitment, no demo deck.