Case study · 75-day engagement
A Miami dermatology practice with two physicians and an aesthetic NP was bidding on $9-CPC keywords like "Botox Miami" and "lip filler near me" — then sending those clicks to a homepage that opened with insurance language and a Medicare-compliant disclaimer. The cosmetic patient could not find the booking button. The medical patient was confused. Both audiences were losing.
The challenge
The owner came in after watching her cost-per-booking creep up for 6 quarters. The bot scored the website at 67 and flagged a fundamental conversion-flow problem: a single landing page for two audiences with opposite intent.
We pulled the data. The damage:
The owner was not running bad ads. She was running good ads sending traffic to a page that did not match the intent. Splitting the funnel was the entire fix.
The plan
We built 11 single-purpose landing pages, one per cosmetic treatment: Botox, Dysport, lip filler, dermal filler full-face, laser hair removal, Morpheus8, microneedling, IPL photofacial, chemical peel, BBL BroadBand, and SkinPen. Each page had treatment-specific imagery (no insurance language, no Medicare disclaimer), pricing transparency, before-and-after gallery, and a one-step booking calendar tied directly to the aesthetic NP's schedule.
Killed broad-match on cosmetic keywords. Built tightly-themed ad groups: each treatment got its own group with exact + phrase match only, and each group routed to its corresponding landing page. CPC dropped 18% the week we shipped this because Quality Score on each ad jumped from 4 to 8.
Insurance-medical patients still needed to convert too. We left the homepage as the medical front door, but stripped it down: hero focused on "Most insurance accepted, same-week appointments for skin checks and acne." No more competing with cosmetic content. Conversion on medical actually improved (from 2.3% to 3.1%) because the page got more focused.
Anyone who landed on a Botox page got Botox-specific Meta retargeting for 14 days, then dropped. Anyone who landed on Morpheus8 got Morpheus8 retargeting. CTR on retargeting jumped from 0.7% to 2.1% because the creative finally matched the visited page.
AI tools we used
We do not pretend the work happens by hand. Three of our internal tools are the reason this engagement moved as fast as it did. Each of them replaces what used to be days of human time.
Custom AI we built that grades every search term on commercial intent, urgency, and local match in under a second. Replaces a half-day human audit.
Takes one bad search term and proposes the related variants we should also negate before they show up. Found 87 negatives nobody had thought of.
Crawls the site and fires every conversion event in a sandbox so you can see what is actually being recorded vs what should be. Caught the duplicate form fire in 4 minutes.
The execution
| Week | What we shipped | What moved |
|---|---|---|
| Week 1 | 11 treatment landing pages designed, copy approved | First 4 pages live by Friday |
| Week 2 | Remaining 7 pages live, ad groups restructured | CPC down 18%, Quality Score up across the board |
| Week 3 | Medical homepage simplified, booking calendars split | Cosmetic bookings up 80% week-over-week |
| Week 4 | Treatment-aware retargeting live, 14-day rolling windows | Retargeting CTR 3× baseline |
| Week 6 | Booked-not-shown reduction: SMS + email confirms tied to each calendar | No-show rate 18% → 7% |
| Week 9 | Pricing-transparency added to top 3 highest-volume pages | Conversion bumps another 22% on those pages |
| Week 11 | Quarterly review | Cosmetic run rate stable at 58/mo |
The numbers
The split funnel produced wins on both sides:
At an average cosmetic ticket of $620 (mostly Botox, with some filler upsell), the +44 monthly bookings represent roughly $27,000/mo in new gross revenue. The retainer pays for itself in the first 4 days of any month.
I had been told for two years that my conversion problem was a website redesign. Mark looked at the funnel for 20 minutes and said: stop sending cosmetic clicks to the medical page. That was the entire problem. Eleven landing pages later, I have a different practice.
Keep reading
Thirty minutes. We pull up your account on screen. We tell you what we would do, what we would not do, and whether we are the right fit for your business.