Google Ads in Healthcare: Restrictions, and Lessons from Million-Euro Budgets

I manage Google Ads budgets at the level of a million euros per year for healthcare brands. At that scale, mistakes scale too: one wrong conversion definition can quietly fund months of waste that looks like success on the dashboard. This article collects the ways healthcare advertising differs from other verticals — and the lessons that grow with the budget.

First, the rulebook: what is allowed in healthcare?

Google's healthcare policies are felt in three areas:

• Personalisation limits: health status counts as sensitive data; remarketing and similar-audience targeting cannot be used in sensitive health categories. The "let me follow everyone who visited my site" playbook does not work in healthcare.
• Content limits: certain treatments and claims (guarantee language, before/after promises) get rejected; some areas require certification or country-level permission.
• Moderation: health ads get reviewed more often. Copy compliance and landing-page consistency are the insurance of continuous delivery.

The practical consequence: in healthcare you build targeting on intent, not identity. The right keyword, the right match type and the right landing page more than fill the space left by restricted personalisation.

Campaign architecture: complexity is a cost

The most common waste I see is accounts that grew in layers over the years: dozens of campaigns, overlapping keywords, nobody sure what drives what. My structure is deliberately simple:

1. Treatment/service-based campaigns: each major treatment gets its own campaign and budget — so budget decisions can follow treatment profitability.
2. Market/language separation: in multi-market accounts, each country-language pair is its own campaign; stuffing ten countries into one campaign makes optimisation impossible.
3. Brand kept separate: brand searches are cheap and high-converting; mixed into generic campaigns they make everything look good and hide the truth.
4. Negative-list discipline: a weekly search-term review — every lira spent on irrelevant queries is a lira that never reaches a patient.

Conversion definitions: trust the patient, not the panel

In healthcare there is a long distance between a submitted form and an actual patient. An account optimised only for form conversions drifts, over time, toward audiences that fill forms but never become patients — because machine learning multiplies whatever you call a conversion. The fix:

• Feed qualified CRM stages (reached, medically eligible, quote sent) back as conversions.
• Include phone and WhatsApp contacts in measurement; a large share of patients skip forms entirely.
• Set cost-per-patient targets per treatment; a blended "cost per lead" average misleads.

The only correct question in a budget meeting: "How many patients did each campaign deliver this month?" Accounts that count forms grow; accounts that count patients win.

Lessons that come with scale

• Scale amplifies errors: a mis-defined conversion invisible at small budgets becomes serious monthly waste at large ones. Measurement first, scale second.
• Don't trust automation blindly: smart bidding works brilliantly with clean data — and brilliantly wrong with dirty data.
• Plan for seasonality: healthcare demand moves with holidays and seasons; splitting budget into twelve equal slices means missing demand.
• No panic on rejections: copy variants, policy alignment and appeals where justified — continuity depends on correction speed.
• Skip the bid-inflation war: everyone loses the click war; the difference is made by landing-page quality and post-conversion speed.

Conclusion

Google Ads in healthcare is a game with learnable rules: policy-compliant structure, simple architecture, measurement wired to actual patients and weekly hygiene. An account built this way spends more intelligently as the budget grows — one without it wastes the same money slightly more expensively every month.