When an agency says "we optimise for bookings", it should mean one specific, technical thing: the ad platforms receive a conversion signal when an enquiry actually becomes a booked appointment, and the campaigns are instructed to find more people like the ones who booked. If the only signal going back to Meta or Google is "someone filled in a form", the campaigns are optimising for leads, whatever the agency calls it.
You can verify which one you are getting with five questions, listed at the end of this article. None of them require technical knowledge, and the way an agency answers them tells you almost everything about how your budget is being spent.
Why does the difference matter so much?
Ad platforms are literal. They deliver more of whatever you count as a conversion. Count form fills, and the algorithm hunts for people who fill in forms, including serial enquirers, price shoppers and people who will never pick up the phone. Count booked appointments, and it hunts for people who behave like your actual patients.
The gap between the two is not academic. Here is an honest range from clinic campaigns generally: out of 100 raw leads from a lead-generation campaign, somewhere between 20 and 50 typically turn into booked appointments, depending on treatment, speed of follow-up and lead quality. Two campaigns can produce the same cost per lead and wildly different costs per booking. If nobody measures past the lead, nobody knows which campaign is the good one, and budget flows to the wrong place by default.
How does the booking signal get back to the platforms?
Bookings happen off the website: on the phone, at reception, in a practice management system. So the platforms cannot see them by themselves. The mechanics are called offline conversions (Meta) or enhanced/imported conversions (Google): the clinic's booking outcomes are matched back to the original ad clicks and uploaded to the platform, which then learns from real bookings rather than form fills.
A working setup needs three things:
- A record of every lead with its source, kept somewhere structured (a CRM or even a disciplined spreadsheet).
- An outcome for every lead: booked, attended, not reached, not suitable. Someone at the clinic has to maintain this; no agency can invent it.
- A regular upload of those outcomes back into the ad platforms, typically weekly.
If any of the three is missing, "we optimise for bookings" is a phrase, not a process.
Worth noting: this is more work than standard campaign management, for the agency and for the clinic. That is precisely why it is rare, and why it separates providers. It is also why an agency doing it properly will ask you for things: outcome data, a named person at reception, a consistent process. Demands like these are a good sign, not friction to negotiate away.
What do agencies usually mean instead?
In practice, the phrase often covers softer things: "we write ads that attract serious people" (creative choices, not optimisation), "we look at your feedback about lead quality" (useful, but anecdotal), or "we optimise for a thank-you page after the booking form" (better than nothing, but online booking forms are a fraction of clinic bookings; most happen on the phone).
None of these are dishonest, but they are not conversion optimisation on bookings. The distinction matters when you compare proposals, because the phrase costs nothing to say.
The five verification questions
Ask these in one email or one call:
- "Which exact conversion event do the campaigns optimise towards?" Good answer: a named event tied to bookings or qualified appointments. Weak answer: "leads" or "engagement".
- "How does a booking that happens on the phone get back into Meta or Google?" Good answer: a description of the offline conversion loop, including what they need from your front desk. Weak answer: silence, or "the pixel tracks everything" (it cannot track phone bookings).
- "What do you need from us for this to work?" Good answer: a concrete list, because the loop always needs clinic-side input. An agency that needs nothing from you is not closing the loop.
- "Show me last month's cost per booked appointment, not cost per lead." Good answer: a number with its calculation. Weak answer: a report full of impressions, clicks and cost per lead only.
- "What happens to leads that never answer the phone?" Good answer: they are recorded and used as negative signal or excluded. Weak answer: nothing; they still count as results.
You do not need perfect answers to all five. You need to see that the loop exists, that someone maintains it, and that reporting reflects bookings rather than stopping at the form.
What if the answers are weak?
Weak answers do not automatically mean a bad agency; plenty of competent teams simply have never built the loop because no client asked. But it does mean your campaigns are optimising on a proxy, and that the reported numbers flatter reality. Raise it directly and watch the reaction: a good partner will treat the question as an opportunity, not an accusation.
If you would like an independent view of what your campaigns currently optimise for and what your real cost per booking looks like, you can request a free audit at brand-developer.com. You get plain-English findings you can act on, whoever runs your ads.