Google Ads accounts for medical practices age faster than for any other industry. The structure that makes them resilient against policy changes, attribution shifts, and reporting evolution looks meaningfully different from the default.
If you have run Google Ads for a medical practice for more than two years, you have lived through at least three significant changes that affected how campaigns perform. Tracking restrictions tightened. Bidding strategies were renamed or absorbed into automated alternatives. Match type behavior shifted. Restricted category policies expanded. Each change altered which campaigns worked and which ones quietly started underperforming.
Most accounts were not built with this volatility in mind. They were built quickly, by a marketing manager or an agency who structured them the way they had structured the last account, with a few campaigns named something generic and ad groups thrown in below. The account ran. It produced leads. And every time Google changed something, the account got a little less efficient, until somebody eventually rebuilt it from scratch.
A durable account structure is not exotic. It is the result of a small number of deliberate decisions made up front, with future change anticipated. Practices that adopt this structure spend less on rebuilds and recover faster from platform shifts.
Start From Service Lines, Not Keywords
The first structural decision is what each campaign actually represents. The most common mistake is organizing campaigns around keyword themes — a “branded” campaign, a “competitor” campaign, a “generic” campaign, and so on. This structure made sense a decade ago. It makes much less sense now.
A more durable approach organizes campaigns around the practice’s actual service lines. Each service line — joint replacement, sports medicine, spine, and so on for an orthopedic group; injectables, body contouring, laser, and so on for a med spa — gets its own campaign. Within each campaign, ad groups segment by intent or sub-procedure. Within each ad group, keywords cluster tightly around a single concept.
This structure produces three benefits that compound over time. It makes the account legible to anyone who looks at it, including the next person to manage it. It allows budget to be allocated by clinical priority rather than by keyword theme, which is how practice owners actually think about their business. And it survives policy changes more cleanly, because the categorical organization mirrors the way Google’s own restricted-category policies are scoped.
Naming Conventions Matter More Than They Seem
A campaign called “Brand Campaign 2025” tells you nothing useful eighteen months later. A campaign called “ORTHO_Knee_Replacement_Branded_2026” tells you the service line, the sub-service, the intent layer, and the year — all from the name.
A working naming convention has four components. The service line (or vertical, for multi-specialty practices). The sub-service or condition. The intent layer — branded, generic, competitor, condition-based. And a temporal marker that lets you distinguish campaigns from different periods without losing the historical data.
Naming conventions sound clerical. They are not. They are the difference between an account that can be audited, optimized, and handed off cleanly and an account that requires institutional memory to navigate. Practices that adopt a convention from the start save substantial time later. Practices that retrofit one onto an existing account often find it is one of the highest-value cleanup activities they have done.
Separate Brand From Non-Brand
Branded search — patients searching for the practice by name — and non-branded search — patients searching by service or condition — behave so differently that mixing them in the same campaign distorts every metric. Branded conversion rates are dramatically higher. Branded cost per acquired patient is dramatically lower. Mixing them produces averages that mislead.
Branded campaigns should always sit in their own campaign, ideally with their own budget and bidding strategy. This keeps non-brand campaign metrics honest and prevents the optimization algorithm from confusing the two intent profiles.
It also protects against a common failure pattern. When a brand campaign and a non-brand campaign share budget, the algorithm — left to its own preferences — will pour spend into the brand campaign because the brand campaign converts better. The practice ends up paying Google for visits that would have happened organically anyway, while the non-brand campaigns starve. Separating brand from non-brand prevents this drift.
Match Type Discipline
Match types in Google Ads have changed meaningfully over the past several years. Broad match has become more aggressive. Phrase match behavior has shifted. Exact match is less exact than it used to be.
For healthcare, where the cost of irrelevant traffic is high and the regulatory risk of certain queries is real, match type discipline matters more than in most industries. A practice running broad match without rigorous negative keyword management will inevitably attract traffic from queries that are not just unproductive but actively unsuitable — wrong specialty, wrong intent, wrong geography.
A durable account structure typically starts with phrase and exact match for established keywords and uses broad match selectively, in dedicated ad groups, with aggressive negative keyword lists in place. This costs some volume in the short term and produces dramatically better lead quality over time.
Build a Negative Keyword Architecture
Negative keywords are the most underused tool in healthcare Google Ads. A well-built account has multiple layers of negatives — account-level negatives for terms that should never trigger ads, campaign-level negatives for terms irrelevant to that service line, and ad-group-level negatives for terms that belong to a different ad group within the same campaign.
Healthcare-specific negative lists should include: jobs and employment queries unless the practice is hiring, free-service queries unless the practice offers free consultations, condition queries the practice does not treat, demographic terms the practice cannot legally exclude in targeting but can exclude in keywords (with care), and any terms that have produced low-quality traffic in past data.
Negative lists are not set-and-forget. They should be reviewed monthly against the search terms report, with new negatives added based on what is actually being searched. Practices that maintain this discipline see meaningfully better lead quality than practices that built a negative list once and never updated it.
Geographic Targeting Architecture
Single-location practices have it easier than multi-location practices, but both benefit from deliberate geographic structure.
For single-location practices, the question is the targeting radius. The default radius — often a generic city-wide setting — is rarely optimal. A more durable structure defines the radius based on actual patient draw, which is usually visible in the practice’s own scheduling data. Patients beyond that radius rarely convert, and bidding on impressions from outside it wastes budget.
For multi-location practices, geographic structure is more consequential. Each location should typically have its own campaign or campaign group, with targeting scoped to that location’s catchment area, separate budget, and location-specific landing pages. Combining multiple locations into one campaign with broad targeting almost always underperforms location-specific structure.
Bid adjustments by location — increasing bids for the most valuable catchment areas, decreasing them for marginal areas — provide an additional layer of optimization that most accounts ignore.
Conversion Tracking That Survives Restructuring
Account structures get rebuilt. Conversion tracking should not have to be rebuilt every time.
A durable conversion tracking architecture separates the tracking layer from the campaign layer. Conversions are defined at the account level, tied to specific events on the website or CRM, and assigned to campaigns through tagging rather than embedded in campaign settings. This way, when a campaign is rebuilt or paused, the conversion data persists. The historical record is preserved.
For healthcare specifically, the conversion tracking should also align with the HIPAA-conscious architecture described in Cluster 2 — server-side, BAA-covered where applicable, with sensitive parameters stripped before transmission. A durable account structure assumes this infrastructure is in place; it does not need to be redesigned every time the campaigns are reorganized.
Asset and Extension Management
Sitelinks, callouts, structured snippets, location extensions, and call extensions all contribute to ad performance and visibility. They also drift out of date faster than most accounts realize.
A durable structure manages assets at the account level where possible, with extensions inheriting to campaigns and ad groups rather than being recreated for each. Out-of-date callouts referencing seasonal offers from two years ago, sitelinks pointing to pages that no longer exist, and location extensions for closed offices are common in older accounts. A monthly extension audit catches these before they accumulate.
Documentation
The most overlooked element of a durable account structure is the documentation that explains it. A short document — accessible to anyone who manages the account — that describes the naming convention, the rationale for campaign segmentation, the conversion definitions, the negative keyword strategy, and any non-obvious decisions makes the account legible to the next person.
Without this documentation, every change in agency, in marketing personnel, or in leadership produces a learning curve and often a partial rebuild. With it, transitions happen cleanly and accumulated knowledge persists. The documentation takes an hour to write and saves dozens of hours over the account’s life.
Why Durability Matters
The reason all of this matters is that healthcare Google Ads accounts go through structural changes more often than most marketers expect. Policy updates restrict categories. Bidding strategies change. Account-level features get introduced or deprecated. Each event tests the resilience of the underlying structure.
An account built durably absorbs these shocks with adjustments. An account built without structure usually requires a rebuild every two to three years, with the loss of historical data and learning that comes with it. The durable structure is not more work to set up. It is just more deliberate. And it pays back every time the platform changes — which, in healthcare advertising, is constantly.
