In Line Dentistry
Performance
What the advertising spend produced, and which parts of the answer are not yet measurable.
Snapshot
Read 2026-09-02
Ad spend
$1,904
Meta since Aug 11 · Google last 30 days
Cost per Meta lead
$10.46
51 lead-form submissions
Cost per Google conversion
$114
calls + booking requests — not the same unit as a form fill
Cost per booked appointment
$1,904
1 booked across both channels
Spend by campaign
Lead to chair
- Meta lead submitted 51Windsor facebook_leads, form IL-IMPLANT-1990-AUG.
- Reached the CRM 49Contacts with source Facebook. The integration works — nothing meaningful is lost here.
- Automated first message 46Tagged speed-to-lead-sent. Sent by an automation built in the GHL UI, not by this repo's specs.
- Any human follow-up 14Carries any status, engagement or callback tag beyond the automated send. The other 35 got one text and nothing since.
- Booked 1Tagged st-booked.
- Showed not trackedNo appointment table is exposed by the GoHighLevel connector. Attendance cannot be read from here.
- Collected not trackedGoHighLevel holds no transaction or invoice rows for this location. Collections are reported manually by the practice.
Recorded pipeline & collections
Opportunity value in the CRM
$13,930
Across 14 opportunities for the year, 12 of which carry no value at all. Both valued opportunities came from Paid Social.
Collections
not tracked
No transaction or invoice rows exist in GoHighLevel for this location. The practice reports collections manually.
Data integrity
Read these before trusting any figure above. Each is a known defect in the source system, not in this view.
- critical Replies land and nothing routes them — W2 is the broken pieceRESOLVED 2026-09-02, and not the way this file first recorded it. Every Meta lead created up to Sep 1 14:40 carries speed-to-lead-sent and none created after does. That is the DUPLICATE-PATH DISABLE LANDING, not a failure: speed-to-lead-sent is the old Facebook path's tag, and switching that path off was the Sep 1 ruling. The W0/W1 marker st-new is still applied to every new lead, and W0/W1 is demonstrably sending — T2 went out a day after entry for two leads, and four more received an Isabel follow-up on the morning of Sep 2. Delivery works and the nurture works. What does not work is W2: replies arrive and nothing routes them. The distinction matters because "leads are not arriving" and "replies are not handled" have completely different fixes, and only the second is true.
- critical Closed-Lost forecasts higher than Closed-WonThe Website Leads pipeline still runs GoHighLevel's default linear probability spread: Closed - Won at 75%, Closed - Lost at 87.5%. A dead deal forecasts higher than a won one, and GHL applies these itself. Documented in ghl-catalog.ts on 2026-08-27 with the fix named — two edits in the GHL UI, Won to 100 and Lost to 0. Confirmed still live 2026-09-02.
- critical 12 of 14 opportunities carry no valueFourteen opportunities exist for the whole year; twelve read $0 and none has ever closed won. Only two carry money ($3,980 and $9,950), both Paid Social. Every pipeline-value figure downstream inherits this.
- warning Collections are not instrumentedGoHighLevel holds no money data for this location, so true ROI and the $30k collections threshold cannot be computed from connected systems.