A patient searching for a dermatologist who wants a Medicare rebate has to visit a GP first. Days or weeks later a referral arrives from a practice you never advertised to. That is not an attribution failure, it is how the system works, and the tracking has to be designed around it.
For specialist dermatology the sequence runs: search, then a GP appointment, then a referral written and sent, then your practice books the patient, then they attend. The Google attributable window covers the first step and nothing else. Everything between the click and the referral happens inside a different practice's systems.
So the honest architecture has two tiers. A micro conversion you can actually fire on your own site, which is an enquiry, a call, or a referral information download, and which the bidding can use. And an offline import of attended appointments, which is the truth and which is reported rather than bid on.
Anyone reporting cost per booking on a specialist dermatology account inside a 30 day window is measuring the wrong thing, and usually reporting a number far worse than reality because most of the conversions have not arrived yet.
A booking-only call to action fails most qualified specialist dermatology traffic, because the reader cannot book yet. A referral request page, a what to ask your GP explainer, or a form that emails them a referral request to take in, all give the visitor an action that matches where they actually are.
The click identifier goes into your own lead store when the enquiry or download happens, because that is the last point at which you control the page.
Bp Premier can export the appointment book to CSV, and critically the export carries a Booked by column that lets you filter to online bookings specifically. That is the closest thing to a native attribution field in the Australian primary care stack.
By the time the patient attends, the original click identifier is usually long gone or was never captured, because the referral came through a channel you did not control. Enhanced conversions matching on hashed email is the realistic route.
Show the distribution, not just the average. A specialist dermatology account with a median 5 week click to attendance gap will look broken at 30 days and healthy at 90, and the client needs to see that before month one rather than after.
| System | Used by | Online booking | Data out |
|---|---|---|---|
| Bp Premier | GP-led skin cancer clinics | Best Health Booking | Appointment CSV export including a Booked by column. API via Halo Connect, partner gated |
| Genie and Gentu | Specialist dermatologists, majority share | None native. Partner integrations only | FHIR and OAuth partner API, closed beta, $500 per month or 20 per cent of revenue |
| Zedmed | GP and specialist | Yes | Halo Connect, doubly gated. The only system listing a dermatology imaging integration |
| HotDoc | Booking layer over the above | Yes, this is the product | No public developer API found |
| Automed | GP booking layer writing into Bp and MedicalDirector | Yes | The only vendor documenting UTM source tracking and analytics reporting |
Since 31 December 2025 all third party integrators to Bp Premier have been required to use Halo Connect and be approved as a Bp Partner, which changes the timeline on any integration project.
A senior team that lives inside your account. Conversion infrastructure wired in before any ad goes live. Reporting that ties spend to booked appointments, not impressions.
A senior specialist who runs clinic accounts day to day.
Every keyword, headline and account structure built inside real Australian clinic accounts.
Winner, Best Google Ads Campaign, Australia 2025. Google Partner. AFR Fast 100.