| Job | Branch | BU | Customer | Job type | Canceled | Days | By | Reason | Outcome | Memo |
|---|
cx_jobs Peaka dataset; all tiles aggregate client-side from the job-grain fact table. Cancel rate = cancels (by cancel date) ÷ jobs booked (by created date) in the selected window — a period ratio, not a cohort rate. When "Entry error / test" is hidden (default), tiles and charts exclude duplicates/test records; the job detail table and the Excel export always contain everything. Rebooked/Transferred are capped at 30 days after the cancel. Branch rates on fewer than 40 booked jobs are grayed as low-volume. Booked excludes marketing-only pseudo-BUs. Job numbers deep-link to the correct ServiceTitan tenant per brand. Rebooked = same customer, same tenant, later non-canceled job created after the cancel. Transferred = same customer name matched across ASV↔MRE tenants (name match understates true transfers). Outcomes and dispatch stage exist on job-level rows only. After dispatch = a technician visit was logged on the job (assignment reached Done) before the cancel — a floor, since mid-route cancels may not log a visit. Entry error / test = duplicates and test records. Memos truncated at 140 chars. Data: ServiceTitan via Peaka, trailing 60 days, rebuilt hourly (:07 CT).