Channel 04 · Attribution
Reported in calls. Not impressions.
An impression never fixed a furnace. LocationCast puts a tracked phone number on every location and every channel, so the monthly question — is it working? — gets answered in the only unit a service business banks: the phone ringing, per location, counted by the week from the call records themselves.
How the counting works
A tracked number for every location and channel
Tracked numbers, assigned per location and channel
Each location publishes its own tracked number, and each of that location’s channels carries it, ringing through to the location’s real line. A call to that number is a call to that location — counted per brand, per week, from the call records, never estimated. Quote requests sent through the location’s own site form are counted the same way: per brand, per week. The customer experience is unchanged: a local number, answered by your people. And the limit, stated on the mechanism’s own page: no per-piece attribution exists today — nothing can credit an individual short with an individual call, and this report does not pretend otherwise. Where a surface one day carries its own tracked number, the piece that carried it could earn its own credit; engineering that across every surface is a design direction the pilot is scoping, not a mechanism being sold here.
A ledger, not a screenshot
Calls accrue into a per-location ledger alongside what shipped that week: pieces published, offers in force, exceptions raised. The report is computed from records — call logs and production logs — not assembled from platform analytics screenshots the night before the meeting.
Booked-job proxies, labeled as proxies
Where your booking flow allows a marker (a call past a duration threshold, a form tied to a tracked source), the report shows it as a proxy, named as one. A proxy is never silently promoted to a booking in our reporting; the day your dispatch system can confirm bookings directly, the proxy retires.
The dark-location alarm
A location whose channel stops publishing, whose latest batch stalls, or whose tracked calls sit at zero past its seasonal pattern does not wait for the monthly report to be noticed. It raises an exception to a human operator the week it happens — that is the difference between a dashboard and a watcher.
The monthly operator report
One page per location. One page for the group.
| Section | What it shows | Source of record |
|---|---|---|
| Calls by location | Tracked calls per location for the month, with the prior month and the season baseline beside them. | Call logs |
| On air while it rang | Each location’s tracked calls printed beside what that location had live that week — flagship, shorts, TV, profiles — so a ringing location is visible next to its programming. Granularity is brand-week: the report credits locations and weeks, not individual pieces, because per-piece attribution does not exist today. | Call logs × production log |
| Velocity | Pieces planned vs. pieces shipped, per location. A location that was promised four episodes and got four says so; one that got two says that too. | Production log |
| Exceptions | Every failure, bounce, stall, and dark period in the month, with what was done about it. This section is never empty by editing. | Exception log |
| Facts currency | When each location’s facts file was last refreshed, and which offers expired mid-month and rotated out on schedule. | Facts files |
| Group page | Locations ranked by calls, the group’s dark list (today: empty is the goal), and next month’s seasonal notes per trade. | All of the above |
What you will not find in it
Impressions as a headline. Follower counts as an achievement. Averages that hide a dead location inside a healthy group. Revenue figures on shared dashboards — money stays behind a PIN by design in this studio’s systems, because reports get screenshotted in truck cabs. And no metric we cannot trace to a record.
The failure doctrine
Every stage of the system must prove its own effect, on a schedule, or report red. A watcher that has gone silent is itself an exception. This doctrine costs us comfort and buys the only thing measurement is for: when the report says a location is ringing, you can dispatch against it.
What the pilot is measuring right now
Baselines before claims
The pilot’s first job is a baseline: per location, the pre-pilot state of presence and tracked calls, established before cadence begins. Against that baseline, the January gate asks whether content shipped on schedule, whether locations stayed lit, and whether tracked-call activity moved at a majority of locations. The gate criteria are published on channel 06, and the verdict will be read from full records — including the locations where nothing moved, if that is what the records say.
Until then, this page describes a measurement system, not measurements. Any vendor who shows you results before their baseline exists is showing you a story.