Part of the On Demand Medical Group series · one client, three lenses
This is the Performance Marketing. On Demand Medical Group comes apart into three deep-dives — same client, three disciplines, readable in any order: overview, Funnel & CRO, and Data & Analytics.
We work with a leading on-demand doctor & wellness brand— an on-demand doctor service and an IV-wellness service under one operation. This page is the Performance Marketing lens: the two demand engines and how each earns its return.
The old approach treated media as one lever: spend more to grow more. We split it into two engines with two jobs — a Google core that harvests the demand already searching, and a Click-to-WhatsApp engine that creates new demand at the top of the funnel. Kept separate and measured separately, the blended return rose from 3× to >6× on <20% less spend.
The principles we rebuilt the media on.
- Harvest and create are two different jobs. Search can only capture demand that already exists. Creating new demand is a separate job for a separate channel — so we stopped asking one engine to do both.
- Concentrate budget on what converts. A tight target was starving demand and the core tier converts cheapest. We lifted the target where it suppressed volume and boosted the proven converting terms — efficiency and volume, not one at the expense of the other.
- Test before you commit. The new channel ran a two-week test before it got real budget. It held cost per conversation below the core-engine benchmark — proof first, scale second.
- Bid per service, per product. The two services convert differently, so they're bid differently — the IV-wellness service on its drips, the on-demand doctor service on its urgent-need terms — instead of one blended setting for both.
Two engines, two jobs.
| The engine | Its job | What we did — and what it returned |
|---|---|---|
| Googlethe efficient core | Harvest the high-intent demand already searching | Rebuilt into location tiers and per-product campaigns, recalibrated bids, boosted converting terms — cost per click −28% and cost per conversation −26% in the eight days after, volume holding. |
| Meta · CTWAthe reach engine | Create new demand at the top of the funnel | Launched Click-to-WhatsApp, ran a two-week test, held cost per conversation below the Google benchmark — proven, now scaling to a full-funnel campaign set. |
| Creativethe fuel | Feed the reach engine what converts | Retired the high-cost creatives and scaled the symptom-led and product-led statics that were winning — then built them onto a full-funnel structure (top, middle, bottom). |
The shift, before → after.
| Dimension | Before | After |
|---|---|---|
| Media model | One lever — spend more to grow more | Two engines, two jobs, measured separately |
| Google core | Blended setup, tight target starving demand | Location tiers and per-product bids, budget on what converts |
| Demand creation | Search asked to create demand it can't reach | Click-to-WhatsApp creating new demand, tested then scaled |
| Creative | High-cost creatives left running | Winning statics scaled onto a full-funnel structure |
| Return & spend | 3× blended return | >6× blended return on <20% less spend |
Results — the media engine.
Return as a multiple; everything else relative, month on month or within-period.
Honest read: the reach engine is three weeks old — its in-test return is read as early signal, not a settled number, and it's now scaling into a full-funnel set. The efficiency win on the core engine is a within-period result: the eight days after the change versus the prior three weeks. The blended return is the trustworthy headline because per-channel tagging still covers only the tracked minority — fuller tagging next period brings the per-service view up to the whole picture.
What the rebuild taught us about paid media.
- Don't ask one channel to do two jobs. — Search harvests, social creates. Split them, measure each on its own job, and both get better.
- A tight target can starve demand. — Over-tight bidding suppresses the very volume it's trying to make efficient. Loosen where it's choking delivery, concentrate where it converts.
- Test before budget, always. — A two-week proof at benchmark cost is what earns a channel its scale — not a hunch and a big spend.
- Two services, two bids. — Different demand converts differently. Bidding each brand and product on its own terms beats one blended setting for both.
Two engines, two jobs — and a return that doubled on less spend .
Google harvests the demand that's already searching at a cost that keeps falling; Click-to-WhatsApp creates the demand search can't reach, proven and now scaling. The next job is the funnel that converts them — and the measurement that reads it all.
Book a strategy call →The rest of the On-Demand Medical Group engagement.
On-Demand Medical Group was one healthcare engagement built across urgent-care and recovery demand. Keep going — here’s the rest of the set:
Overview — one confidential medical group, two demand categories, and multiple booking journeys built under one clearer performance system.
Performance Marketing — separated broad medical demand into clearer service-led funnels, from traveler illness to recovery care. More qualified inquiries from people with stronger booking intent.
Funnel & CRO — rebuilt the path from ad click to WhatsApp booking so each service had a clearer offer, clearer next step, and less friction before conversion.
Data & Analytics — tracked performance by service line, audience intent, location, and lead quality, turning mixed healthcare demand into a clearer picture of what actually drove bookings.