Measurement work, told end to end
How I connect ad spend to real outcomes: offline conversion loops, attribution rebuilds, and the booking-system bug that was quietly costing patients. Client details are removed, the engineering is not.
Prefer writing? (I reply within 24 hours)Prefer escrow? Hire me on Upwork (escrow)
The services behind these case studies:Website + BookingBooking + TrackingPatient Conversion TrackingGoogle Consent ModeMarketing Attribution
Before Moving the CRM, I Counted It: 0.4% Had an Email
A migration planned from counts, not from what the dashboard implied
The plan assumed big newsletter lists. Counting the CRM through its API showed 0.4% of contacts had an email, so the migration plan changed.
Five Tracking Failures Nobody Noticed
Tracking that runs only with consent, one pixel, no lost form submissions
A cookie banner that had to mean no, two pixels, a fake €1,140 order, analytics with no keys and a form that lost requests: five fixes, measured.
Sales Stopped Overnight: Was It the Website or the Ads?
Root cause the same day, with evidence for the ad agency
A client wrote that orders had stopped. The same day I showed the website worked and traced the drop to one Meta campaign that had switched off.
US medical practicesThe HIPAA Boundary in a Marketing Data Pipeline
A boundary you can show a compliance officer on one page
Where patient data stops and ad tech begins, drawn as an architecture: two zones, a scrub before load, and a table that physically cannot hold PHI.
Multi-source marketing measurementThe Marketing Pipeline That Re-Reads Yesterday
Reports that stop changing after you send them
Ad platforms restate history and CRMs edit rows late. The ingestion design that survives it: watermarks, per-source lookback, idempotent loads.
US medical practiceOffline Conversions: Google Ads to Real Patients
Patient visits reported back to Google Ads, not just form fills
I built an offline conversion loop that ties Google Ads spend to patients who actually showed up, using a HIPAA-conscious click-id-only upload.
Multi-location clinicGoogle Ads Attribution Rebuild: Clinic Case Study
Every booking and call keeps its real source, no double counting
How I rebuilt Google Ads attribution for a multi-location medical practice: bookings keep their source, calls carry a source, no double counting.
US medical practiceThe Screen Said 'You're Booked.' Some Patients Weren't.
Matching and success screen fixed, email and text fixes documented
Some patients booked online and saw a confirmation, but the appointment never landed on their chart. Four failures hid behind that one screen.