The Meta Ads playbook we run for hair transplant clinics

Most clinic accounts we audit are running fifteen ad sets, four campaigns, and a Lookalike from a pixel that has fired twice. The account is not the problem. The structure is.

Here is the shape we use, why we use it, and the parts that most teams get wrong.

One CBO campaign, three ad sets, tight creative rotation

We consolidate spend into a single Advantage+ or CBO campaign optimising for the lowest event in the funnel that has enough volume, usually Lead or Complete Registration, never Landing Page View.

Inside that campaign we run three ad sets:

• Broad, no interests, no lookalikes. Meta's model has more signal than any interest stack you can build.
• Retargeting, 30-day site visitors plus 90-day IG/FB engagers.
• Lookalike 1-3% from a seed of high-intent converters. Booked consults, not leads.

Anything beyond that fragments learning and starves ad sets of the fifty conversions per week they need to exit the learning phase.

Creatives do the work; the algorithm does the targeting

For hair transplant, the winning creatives share four traits:

1. Real patient footage. Before/after stills lose to 15-second selfie-style clips from real patients.
2. The founder or a doctor on camera. Trust is the bottleneck; a person on camera collapses it faster than any brand asset.
3. A clear objection handled in the first three seconds. "Will it look natural?" "Can I fly home the next day?" That is the hook.
4. Native aspect ratio and captions. 9:16 for Reels/Stories, 4:5 for Feed. Burned-in captions, always.

We ship six new concepts every two weeks and kill anything below a 1% CTR at £15 spent.

The funnel is where the money actually leaks

The lead ad is easy. Converting the lead into a booked consult is where clinics lose weeks of budget.

• WhatsApp beats email, every market. Response times drop from hours to minutes. Show-rates roughly double.
• First message inside five minutes or you will lose 40% of the leads to whoever replied faster.
• Qualify before you sell. Norwood scale, budget expectation, travel window. A booked consult with a Norwood 2 who wants 4,000 grafts is a wasted slot.

If you cannot staff a five-minute response window, this is exactly where AI follow-ups pay for themselves.

What we ignore

• Interest stacks for anything except retargeting. Broad wins in 9 out of 10 accounts we run.
• Advantage+ Shopping for a service business. Wrong tool.
• CAPI without server-side deduplication. A broken CAPI setup is worse than none; it double-counts and poisons the model.

The number that matters

Not CPL. Not CTR. Booked consultations at target cost per booking. Everything above is a proxy. Everything below is theatre.

If your clinic is running Meta Ads and you cannot pull that number in one screen, that is the first thing to fix, before you touch a single ad set.