Campaign structure by specialty and procedure

A good healthcare PPC campaign structure puts specialties and units at campaign level, where budget and location are set, and procedures or intent clusters at ad group level, each with its own landing page. Keep brand separate, split out high-value procedures only when they need their own budget or geography, and consolidate where data is thin.

Healthcare PPC campaign structure is the least glamorous decision in a hospital Google Ads setup and the one that decides almost everything else. It decides whether the CFO can see what orthopaedics cost last month, whether a cardiology click can be bid differently from a dermatology click, whether each ad lands on the right page, and whether automated bidding has enough data to learn.

I have inherited accounts at both extremes. One had a single campaign called “Hospital” with a few hundred keywords and one landing page. Another had a campaign per doctor per procedure per city, most of which spent almost nothing and learnt nothing. Both were unmanageable for the same reason: the structure did not follow how the hospital makes decisions, or how patients search.

This article is the structure I now start from, with the reasons, the exceptions and a build sequence. It sits within the wider complete guide to Google Ads for doctors in India.

What each level of the account controls

Everything follows from what Google lets you set at each level, which the Google Ads account structure help page lays out clearly:

• Account: billing, access and account-wide settings such as account-level negative keywords.
• Campaign: budget, bidding strategy, locations, languages, networks and ad schedule.
• Ad group: a tightly themed set of keywords, the ads that answer them and the landing page they point to.

So the rule is simple. Split into separate campaigns wherever you need separate budget, bidding, geography or schedule. Split into separate ad groups wherever you need a different message or a different landing page. Everything else is detail.

Google also warns that restructuring later throws away accumulated data. That is one more reason to get this right before launch rather than after the first quarter’s review.

The healthcare PPC campaign structure I start from

For a hospital or a multi-specialty clinic, the default is a campaign per specialty per location, with ad groups per procedure, condition or intent cluster. For a single doctor, the same logic applies one level down: a campaign per service line, ad groups per procedure.

Why specialty at campaign level? Because budgets, capacity and the value of a new patient are set by service line in almost every Indian hospital I know. A unit head allocates OPD slots by department. Finance reviews revenue by department. If the campaign mirrors that, the Google Ads report becomes a management report without a spreadsheet in between. I have written about building budgets this way in a hospital marketing budget built from service lines up.

Here is how one specialty might break down. The names are illustrative, not a template to copy blindly:

Ad groups stay small: a handful of closely related keywords that one ad and one page can answer honestly. If you find yourself writing an ad that has to cover two procedures, you need two ad groups.

Health checks, diagnostics and second opinions usually deserve campaigns of their own even though they cut across specialties. The buyer is different: a health check is often bought by a family member or an employer, a diagnostic test is often prescribed elsewhere, and a second opinion is a patient already under treatment somewhere else. Each needs its own message and page, and none of them sits naturally inside a single department’s campaign. Second opinions in particular are mishandled by most hospitals, a point I make in second-opinion demand.

When a procedure earns its own campaign

Some procedures should break out of the specialty campaign into their own. I look for any one of these signals:

• Budget share: the procedure would otherwise take most of the specialty budget, starving the other ad groups.
• Different geography: patients travel from other cities or states for it, while consultations are local.
• Longer decision cycle: the gap between enquiry and treatment is long enough to need its own conversion setup and patience before judging it.
• Separate capacity: the procedure depends on a specific surgeon or a specific theatre schedule, and ads must pause when that capacity is full.

Joint replacement, cardiac procedures, bariatric surgery and some oncology services often meet these tests. Routine consultations almost never do.

Brand, doctor names and competitors

Brand searches belong in their own campaign, always. They are cheap, they convert well, and if they sit inside a specialty campaign they make that specialty look far better than it is. Separating them also lets you see whether brand demand is growing, which is a useful signal in itself.

Doctor-name searches are a judgement call. For a well-known senior consultant, people search the name directly, and a small ad group in the brand campaign protects that traffic and sends it to the doctor’s profile page. Make sure the page is worth landing on; doctor profile pages that convert covers what it needs.

Bidding on other hospitals’ names is something I avoid in healthcare. It rarely produces good patients, it invites trademark complaints if a name appears in ad text, and it does not sit well with the professional tone patients expect from a hospital.

Multi-unit groups and geography

For groups with several units, location is the second split after specialty. A campaign per specialty per unit keeps each unit head’s numbers clean and lets budgets follow local capacity. Where two units serve overlapping catchments, draw the targeting boundaries deliberately so the two campaigns do not bid against each other in the same auction.

Use the location setting carefully. Google’s default of “presence or interest” also reaches people who show interest in a place without being there; Google’s location targeting help explains both options. For local OPD campaigns I usually switch to “presence”. For medical travel campaigns aimed at patients elsewhere, the opposite logic applies, and those deserve their own campaigns with their own pages.

Tier-2 and tier-3 catchments need their own thought. Patients from surrounding towns often search for the city, not the neighbourhood, and they may need travel and stay information before booking. The digital signals worth reading before expanding are covered in tier-2 city expansion: what the digital footprint tells you first.

Governance matters as much as settings. When units run their own agencies, structures drift apart within a year and nothing can be compared. The operating model for keeping one standard across units is in centralising marketing across hospital units.

Settings that belong at campaign level

Because the campaign is where schedule, bidding and budget live, a few operational decisions should shape how you split campaigns, not just how you configure them:

• Ad schedule follows call handling. If a specialty’s enquiries come mainly by phone and the line is covered only in OPD hours, its campaign should reflect that, while a campaign driving WhatsApp or form enquiries can run later.
• Pausing follows capacity. When a surgeon is on leave or a theatre list is full, you want to pause exactly that demand, not the whole department. If that happens often, the procedure needs its own campaign.
• Bidding follows data maturity. A new specialty with little conversion history and a mature one with steady appointments need different bidding strategies, which is easier when they sit in separate campaigns.
• Budget follows ownership. Whoever approves the budget should be able to find it as a line in the account without adding up ad groups.

None of these are Google best practices in the abstract. They are the practical consequences of how hospitals actually run OPDs and theatres, and they are the questions a unit head will ask in the first review.

Language and regional search

I do not usually split campaigns by language for search. Patients in most Indian cities search in English or in Hindi and regional languages typed in Roman script, and a well-built ad group catches both if the keywords reflect how people actually type. Where there is real volume in a regional script, a separate ad group with ads and a landing page in that language performs better than translated English copy.

The test is the search terms report. If regional or Hinglish queries appear in volume and convert, give them their own ad groups. If they do not, do not build for them speculatively.

Naming conventions that survive agency changes

A naming convention sounds trivial until a new agency inherits the account or a finance analyst tries to join Google Ads data with the CRM. I use a fixed order: channel, unit, specialty, theme, match or targeting note. For example, “Search | City A | Ortho | Knee replacement”. Keep the specialty codes the same as the CRM and HIS use, so reports join without manual mapping.

Labels do the rest. Google Ads lets you attach labels to campaigns, ad groups and ads, and I use them for things that cut across the structure: which campaigns belong to a launch, which ads went through clinical review and when, which ad groups feature a particular doctor. When that doctor leaves or changes OPD days, a label filter finds every affected ad in seconds instead of an afternoon of searching.

Write the convention down, share it with every agency and treat deviations as defects. It is a small discipline that pays off every month.

A build sequence for a new hospital account

This is the order I follow when setting up a hospital Google Ads account from scratch:

1. Map services to decisions. List specialties, priority procedures, units and the person who owns each budget.
2. Check capacity. Confirm which doctors can take new patients and when, so you do not build campaigns you will have to pause.
3. Draft the structure on paper. Campaigns by specialty and unit, ad groups by procedure and intent, one landing page per ad group.
4. Build keyword and negative lists per ad group, plus shared negatives at account level. The guide to keywords that book appointments covers the intent tiers.
5. Build landing pages before the ads, following landing pages for doctor consultation ads.
6. Set up conversion tracking and test every action with a real enquiry.
7. Launch brand plus one or two specialties, then add more only when the first are stable.

The printable version is part of the Google Ads launch checklist for doctors.

When to consolidate instead

Granularity has a cost. Automated bidding learns from conversions, and an ad group or campaign with very few conversions gives it almost nothing to learn from. If most of your campaigns see only a trickle of appointments each month, merge them: combine related procedures into one ad group, or related specialties into one campaign with shared budget, until each has enough signal.

Google Ads also offers shared budgets and portfolio bid strategies, which let several campaigns draw on one budget or one bidding strategy while keeping separate reporting. For a hospital with many thin specialty campaigns, that is often a better compromise than merging campaigns outright, because each department still sees its own line while the bidding learns from the combined pool.

Performance Max and AI Max change the calculation again, because they deliberately work across keywords and pages. They can sit alongside a clean search structure, but they are not a substitute for one. I cover where they fit in Performance Max for healthcare: when it works.

Questions people ask