Build the foundation, then buy the traffic
Onboarding followed the usual sequence. We took access to the practice’s accounts and website, audited the site end to end, and fixed what the audit surfaced — structural issues, alignment and layout problems, and content that needed tightening. Then we created and configured the ad accounts from scratch so performance could actually be attributed rather than guessed at.
The plan was straightforward: paid media drives people to the service pages, they book a programme and pay on the website, and the practitioner runs the sessions online. Clean, self-service, and scalable for a solo practice. It depended on exactly one thing — the ability to take a card payment on the website.
- Why holding the whole stack mattered laterBecause we held the site, the accounts, the tracking and the payment setup together, we could see the blocker coming from the payment side rather than discovering it as a media problem. A media-only vendor would have been told about it weeks later.
A regulated category, and a licence that did not exist
Health and wellness is among the most heavily policed categories in paid media — and in the UAE, in payments as well. We applied for a merchant account with a widely used regional payment gateway. The application did not fail on paperwork. It failed on business classification.
The practice holds a professional licence covering lifestyle consultancy and complementary wellness services. The reviewers’ position was that the website presented its therapeutic modalities in a clinical context, which sits outside that licensed activity. The requirement settled on a different licence class entirely, obtained from a different authority, against a different definition of the practice. We investigated the route thoroughly and it was neither quick nor certain — and the business could not wait on an uncertain licensing outcome to start earning.
- We ran the regulatory correspondence, not the clientAcross those seventeen days the majority of the outbound emails came from our side — submitting revised policies and consent documentation, supplying the licence, proposing three concrete remedies rather than waiting to be told one, and chasing status four separate times as the review moved between the gateway and the bank.
- And when it did not workIt is worth being plain: we did not win this one. The approval never came. What we did do was establish that quickly and definitively enough to stop spending time on it — which turned out to be the valuable outcome.
Seventeen days, and who did the work
| Day | What happened | Driven by |
|---|---|---|
| 2–5 | Website review requirements issued; application submitted, payouts to be suspended until met | Gateway |
| 8 | Updated terms, privacy and cookie policies, consent documentation and pricing pages submitted | Us |
| 8 | Application placed on hold — services presented in a clinical context, outside licensed activity | Bank |
| 9 | Licence submitted with three proposed remedies, and a direct request for the exact requirement | Us |
| 9 | Requirement confirmed: a different professional licence class | Gateway |
| 9 | Regulatory clarification submitted arguing complementary-therapy classification | Client |
| 13–17 | Three further follow-ups chasing status and confirming whether wording changes would suffice | Us |
| 19 | Escalated to an internal call. No approval granted. | — |
Reconstructed from the merchant application correspondence. Third-party names omitted.
Change the funnel, not the paperwork
Every version of the original plan ended at a payment page that could not be built. So we stopped treating that as a problem to solve and started treating it as a fixed constraint, then asked a different question: where else can this transaction happen?
On a phone call, in a conversation. The practitioner already spoke to every client before their first session; the payment could simply be arranged there. That single change removed the gateway from the critical path entirely — the business no longer needed merchant approval in order to trade. It meant conceding that the plan the client had been sold on was not achievable in this category, at this licence class, at this time. Saying that early is uncomfortable. Saying it late is far more expensive.
We also set search aside and moved the entire budget to one channel running click-to-WhatsApp campaigns. Search policy handling of health-adjacent advertising is restrictive, and its strengths depended on the website doing work it could no longer do. At AED 33 a day, splitting across two platforms would have starved both of the conversion signal needed to optimise.
Target the problem, not the demographic
Click-to-WhatsApp is a format, not a strategy. What made it work was matching the ad to the specific problem a person was already carrying, then qualifying them before the practitioner ever spent time on the conversation.
Rather than advertising the modality to a broad wellness audience, we researched and built interest-based targeting around the specific conditions the practice treats, then ran ads mapped to distinct service lines — alongside a general programme for everyone who did not fit neatly into one of them. Someone who has not slept properly in months does not identify as a person seeking hypnotherapy. She identifies as a person who cannot sleep. An ad that names the problem gets recognised; an ad that names the modality has to be decoded first.
We then built qualifying questions into the WhatsApp flow, so people answered a short set of questions the moment the thread opened. A single practitioner cannot triage hundreds of enquiries — volume without filtering is not a benefit to her, it is a second full-time job.
- The number that matters is the last oneMore than half of everyone who tapped the ad opened a real conversation. A website funnel here would have lost most of that traffic before booking — and all of it, since the payment step did not exist.
- The honest boundaryThese are qualified conversations, not bookings. Payment and scheduling happen off-platform.
The funnel, stage by stage
| Stage | Result | Conversion from previous |
|---|---|---|
| People reached | 53,494 | — |
| Impressions served | 254,659 | 4.76 per person |
| Clicks (all) | 4,614 | 1.81% of impressions |
| Tapped through to WhatsApp | 1,275 | 27.6% of all clicks |
| Started a conversation | 668 | 52.4% of link clicks |
195 days, one campaign, one channel, AED 33 per day.
What we built after the ads
Generating 668 conversations for a one-person practice creates a problem as well as an opportunity. Handing over the leads and calling the job done would have been defensible, and useless.
So we set up a CRM and an automated booking system, so enquiries are captured, tracked and scheduled without manual administration. The practitioner does not maintain booking records, chase her own calendar, or rebuild the same information across tools. What is left for her to do is take the payment and issue the intake form. Everything else runs without her.
Lead generation that overwhelms the person receiving it is not a result. For a solo practitioner the constraint on growth is her time, so the highest-value work available was giving her more of it.
We were engaged to run ads. What actually moved the business was a website audit, a seventeen-day merchant dispute, a redesigned sales model, and a CRM build. Not one of those was in the scope of work.