You inherit a healthcare Google Ads account with plenty of clicks, a respectable-looking dashboard, and no reliable answer to the only question that matters: which campaigns produce booked patients? The usual response is to rewrite ads, raise bids, or move everything into a new Performance Max campaign. I don't start there.
I'm Chase McGowan, a self-taught, independent Google Ads and paid media operator. I keep my roster deliberately small, with two to three new clients per quarter maximum, because I want the person speaking with you to be the person fixing the account. In healthcare, that order matters more than cosmetic campaign changes. I audit tracking first, policy exposure second, and account structure third. Only then do I decide what deserves more budget.
Google Ads remains valuable because healthcare searches often carry immediate intent. One 2026 physician benchmark reported an average CPC of $4.76, a 6.61% CTR, a 12.43% conversion rate, and a $40.04 cost per lead, with CPL down nearly 30% year over year. Those figures come from a published benchmark, not a promise for your specialty or market, and the same source makes clear that performance varies by geography and service line (healthcare CPC and conversion benchmarks). Another benchmark places average healthcare search conversion rate at 8.09%, which is why I care more about conversion quality than traffic volume (healthcare search advertising benchmarks).
A typical inherited account has been optimized around activity rather than revenue. Search campaigns run on broad match, condition terms sit beside procedure terms, call conversions count every short interaction, and form submissions pass no meaningful value back to Google Ads. The account may look busy, but the bidding system is learning from weak signals.
The first thing I check is whether the reported conversion is something the practice wants. A consultation request, a booked exam, a qualified call, and a page view aren't interchangeable. If the account treats them as equal, automated bidding will pursue the cheapest event, not the patient most likely to become revenue.
I look for a few problems immediately:
My audit order is simple: tracking integrity, policy exposure, then structural rebuild. Reversing that order creates a cleaner-looking account with the same blind spots.
The ophthalmology account I rebuilt demonstrates why this sequence matters. Platform-reported results showed conversion rate moving from under 1% to 17%+, cost per conversion falling 89%, from about $400 to about $44, and 190 additional patient calls per month. Those numbers don't establish a universal healthcare benchmark. They show what can happen when the account measures the actions that matter and separates intent properly.
The immigration medical exam clinic followed a different pattern. Over two years at approximately $13,000 per month, platform-reported cost per booked exam rose only 10% while market CPCs rose 54%, and conversion rate increased from 11.5% to 15.9%. The lesson isn't that every account can repeat those results. It's that disciplined measurement and controlled structure can protect booked-patient economics when auction pressure rises.
If you want a senior review before changing anything, I offer an Office Hours account review. I'd rather identify whether the problem is measurement, eligibility, intent, or landing-page friction than recommend a rebuild by reflex.
Healthcare advertising isn't governed like ordinary lead generation. Google's policy history shows the direction clearly. In October 2019, Google updated its Healthcare and medicines policy to prohibit ads for speculative and experimental medical treatments globally, after earlier policy changes involving pharmaceutical ads and the .pharmacy top-level domain (Google Ads healthcare policy history). The practical consequence is that claims, destinations, categories, and targeting all need review before launch.
Google treats health as a sensitive interest category in personalized advertising. That category includes physical and mental health conditions, chronic conditions, products or services managing chronic conditions, intimate-body-part health issues, invasive procedures such as cosmetic surgery or injections, and disabilities (Google's sensitive interest categories). That doesn't mean every healthcare search campaign is forbidden. It means audience construction and personalization require more care than ordinary ecommerce targeting.
Patient-facing campaigns need conservative creative and compliant destinations. Avoid guaranteed outcomes, unsupported superiority, and language that implies Google knows a user's medical condition. Before-and-after framing can create policy and trust problems, especially for cosmetic services.
Clinician-facing campaigns are different. Google clarified in May 2025 that the Health sensitive-interest category doesn't apply the same way to content directed at healthcare professionals in their professional capacity. The Restricted Drug Terms category received the same professional carveout effective July 1, 2025 (Google's healthcare professional policy update). I still separate clinician campaigns by keyword, landing page, and offer. A professional audience shouldn't be mixed into patient acquisition because both groups use medical vocabulary.
Prescription and telemedicine campaigns require category-specific review. Some healthcare content can't be advertised at all, while other categories are restricted by location and certification. Prescription drug promotion is prohibited unless the advertiser satisfies Google's specific certification rules, and eligible campaigns can still serve with limitations (Google healthcare and medicines policy).
| Ad lane | Policy gate | Common disapprovals |
|---|---|---|
| Patient acquisition | Allowed service, compliant copy, compliant destination | Guaranteed results, sensational condition language, unsupported medical claims |
| Clinician-facing | Professional-capacity targeting and appropriate destination | Patient-style personalization, mixed audiences, unclear professional intent |
| Insurance or prescription | Location rules, certification, restricted category review | Launching before certification, noncompliant landing pages, restricted product claims |
I verify eligibility in the account before committing spend. For U.S. health and medical insurance, Google requires advertiser certification, while government advertisers are pre-approved. Dental, vision, and travel health insurance coverage aren't restricted under that specific policy (Google health insurance advertising requirements).
The practical sequence is: confirm the service is allowed, inspect every destination page, complete required certification, then monitor ad status. A campaign can be technically eligible and still receive limited serving. For a working set of audit checklists, the Fre Vault Preview is a free preview of vault audit checklists. If you value direct scrutiny, working with Chase McGowan means the operator reviewing the account is the operator making the changes.
I don't launch a healthcare rebuild until the measurement path is understood. Healthcare sites often collect sensitive information in forms, appointment systems, call recordings, and patient portals. The question isn't whether tracking is useful. It's whether the data sent to Google Ads is necessary, permitted, minimized, and configured without exposing protected health information.
Start with Consent Mode v2, which communicates a visitor's consent choices to Google tags and changes measurement behavior when consent isn't granted. It isn't a HIPAA compliance certificate. Your privacy counsel, compliance team, and vendors still need to determine whether the overall implementation is appropriate.
I prefer event data that describes an action without describing a medical condition. A form submission can be recorded as a conversion, but the form content, free-text symptoms, diagnosis details, appointment notes, and patient identifiers shouldn't be passed into ad platforms. URLs and page titles also need review if they contain condition or treatment information.
For enhanced conversions, an email or phone identifier may be normalized and hashed before transmission, subject to the practice's legal and privacy review. I strip unnecessary fields, avoid sending form answers, and document the exact data flow. Server-side tagging can reduce browser-side exposure, but it doesn't make prohibited data acceptable. Moving sensitive information to a server doesn't remove the underlying compliance responsibility.
Call tracking needs the same discipline. The number displayed to the user can be dynamically inserted for attribution, but the destination must remain the practice. Record the conversion only after defining what counts, such as a connected call of meaningful duration or a call later marked as a booked appointment. Raw call recordings and clinical details don't belong in Google Ads.
One concrete action you can take today is to submit your primary form in a test environment and inspect the conversion event. Confirm that enhanced conversions fire, that the thank-you event passes only a permitted hashed identifier, and that no condition, diagnosis, or free-text field reaches the tag. My Chase McGowan's tracking guide covers the implementation logic, but the account's compliance review remains specific to your practice.
I don't build healthcare accounts around the old single-keyword-ad-group habit. The useful separation is by intent and policy exposure, not by forcing every keyword into a microscopic ad group.
I start with four campaign buckets. Brand captures existing practice demand. Condition captures people describing a problem, such as “knee pain” or “migraine.” Procedure captures more explicit service intent, such as “knee replacement” or “Botox.” Competitor campaigns are isolated because they carry different costs, messaging constraints, and landing-page expectations.
| Intent bucket | Example keywords | Recommended match types | Policy risk | Conversion intent |
|---|---|---|---|---|
| Brand | Practice name, provider name | Exact and phrase | Lower, but destination still matters | Often strong |
| Condition | “knee pain,” “migraine treatment” | Exact and controlled phrase | Higher, because intent can be informational or sensitive | Mixed |
| Procedure | “knee replacement,” “Botox consultation” | Exact first, then tested phrase | Service and claim review required | Usually stronger |
| Competitor | Competitor name plus service | Exact and phrase with exclusions | Trademark and comparison review | Variable |
Exact match gives me control while the account establishes clean query data. Phrase match can expand reach, but I watch the Search Terms report closely. Broad match has a place only when conversion tracking is reliable, negatives are mature, and the service line can tolerate exploratory spend. I don't use broad match to compensate for an unclear offer.
The baseline negative list usually includes jobs, salary, lawsuit, DIY, training, school, free, home remedy, and OTC terms where relevant. I add negatives from actual search terms, not from a generic list alone. A query about insurance billing may be commercially useful for one clinic and irrelevant for another.
Ad groups should contain closely related queries with a shared message and destination. I typically begin with responsive search ads built around that shared intent, then consolidate when the data shows that extra fragmentation adds management overhead without improving control. For multi-location practices, I separate location targeting and local proof while avoiding duplicated campaigns that compete against one another.
Patient language is a diagnostic tool. Patients search with symptoms, fears, locations, and practical constraints. Clinicians search with referral terminology, product names, procedure specifications, and professional context. If both appear in one Search Terms report, that doesn't automatically mean the campaign is broad enough. It may mean the account has no separation between audiences.
For additional ideas outside the account, competitive keyword discovery can help identify how adjacent providers frame services. I use that as research, then validate every term against actual patient intent and policy eligibility. The account itself remains the source of truth. This guide to optimize Google Ads account organization is useful when the current structure has grown by addition rather than design.
Bidding should follow measurement maturity, not platform fashion. I start with Maximize Conversions, which asks Google Ads to seek as many selected conversion actions as possible within the budget, but I don't feed it every form fill and phone interaction. The campaign needs a sensible CPA ceiling in the business plan, even if I don't force a strict target immediately.
Target CPA, or tCPA, tells Google Ads to pursue conversions around a chosen acquisition cost. It can work well for established service lines, but a narrow procedure with limited volume can become under-delivered when the target is too aggressive. I introduce it only after the campaign has 30 or more conversions per month, as a practical operating threshold, not a Google guarantee.
Target ROAS, or tROAS, optimizes toward a return based on conversion value. I won't use it until revenue or booked-patient value is trustworthy. A fabricated value makes the strategy precise in appearance and wrong in practice.
Brand defense usually needs enough budget to capture existing demand, but it shouldn't consume the money required to acquire new patients. Procedure campaigns often deserve priority when booking economics are clear. Condition campaigns can support discovery, but I expect more research behavior and stricter query control. Remarketing should remain conservative because healthcare personalization rules limit how I construct and use audiences.
I don't layer sensitive health inferences onto patient campaigns. I use contextual keyword intent and approved audience settings, then keep clinician-facing searches separate. Google's healthcare professional carveout makes professional targeting a distinct policy lane, not permission to mix patient and clinician data.
Current patients, employees, and recent converters should be excluded from acquisition campaigns where the audience configuration permits it. I also review whether a remarketing list itself could imply a sensitive condition. Past site visitors and hashed first-party data may be usable after proper consent and policy review, but the practice shouldn't assume that every audience list is safe because it was collected directly.
For a deeper explanation of when to move between bidding models, see how to maximize ROI with these bidding strategies. The sequence matters: clean conversion actions, stable data, then automated constraints.
Generic healthcare copy wastes the advantage of search intent. “Compassionate care” and “modern facility” are easy for every competitor to claim. I build responsive search ads around details a patient can act on, such as location, accepted insurance where accurate, consultation availability, and the specific procedure or concern searched.
The copy still needs restraint. I don't promise outcomes, imply guaranteed results, or suggest that Google knows the searcher's diagnosis. Clinical proof can be referenced when it is accurate, substantiated, and presented without sensational framing. The safer approach is to describe credentials, process, candidacy, and next steps rather than guarantee what a patient will experience.
A condition query should land on a useful condition page. A procedure query should land on the relevant procedure page. Brand traffic can go to the homepage when that page clearly supports the next action. Sending every click to the homepage creates unnecessary work for the visitor and makes message matching harder to evaluate.
Extensions should support the service line:
The form-versus-call split should reflect how the service is bought. Urgent services often need a phone conversation. Surgery, fertility, and other high-consideration services may need a consultation form plus a call option. I define one primary conversion event per landing page, then treat supporting interactions as secondary. Tracking every button click as a primary conversion teaches bidding to chase activity instead of completed inquiries.
That distinction is why the ophthalmology result matters. The platform-reported improvement to 17%+ conversion rate and approximately $44 cost per conversion came with a business outcome, 190 additional patient calls per month, rather than a traffic-only story (healthcare advertising policy context). I use that kind of downstream action as the standard, while recognizing that no account should borrow another practice's economics.
Healthcare optimization is a cadence, not a launch event. Weekly work keeps waste from accumulating. Monthly work answers whether the service line deserves more budget, a different destination, or a different commercial expectation.
I review remarketing audiences with a narrow filter. Past site visitors, video viewers, and properly consented Customer Match data using hashed first-party identifiers may be workable, subject to policy and privacy review. I avoid audience definitions that infer a person's medical condition from behavior, and I treat in-market health segments and demographic health inferences as restricted territory rather than default targeting options.
When a service line underperforms, I don't immediately lower bids or pause it. I check match-type bleed first, then landing-page friction, then tracking integrity, then policy eligibility. Only after those checks do I change bidding or budget. A campaign can't bid its way out of a form that fails, a phone number that doesn't connect, or a service Google won't approve.
Enhanced conversions and offline call imports should eventually feed booked appointments back into Google Ads. Smart Bidding then has a chance to optimize toward completed business outcomes rather than raw form fills. The exact data design depends on the practice, but the principle is fixed: the platform needs a clean signal, and the practice must control what information leaves its systems.
| Task | Cadence | Why it matters for healthcare PPC |
|---|---|---|
| Search-term pruning | Weekly | Prevents condition, research, employment, and irrelevant queries from consuming budget |
| Negative keyword updates | Weekly | Keeps match expansion tied to actual patient intent |
| Service-line budget review | Weekly | Protects high-value procedures from flat allocation |
| Impression share analysis | Monthly | Distinguishes budget limits from ranking or ad-quality limits |
| Cost per booked patient | Monthly | Connects ad spend to the outcome the practice actually values |
| Policy and certification check | Monthly and after changes | Reduces disapprovals and limited serving |
| Creative and landing-page review | Monthly | Keeps claims compliant and the message aligned with intent |
My own operating model is deliberately different from a traditional agency. Come Together Media LLC offers a $7,500 30-day Google Ads Sprint, built around tracking first, then account rebuild and optimization. Ongoing management starts at $3,500 per month, month to month, and a $350 Office Hours session is available for a focused account review. I keep the roster small, don't use account managers, and don't hand the work to juniors.
If your healthcare account is spending real money but can't connect Google Ads to booked patients, hire me for 30 days before you hire me forever through the Google Ads Sprint. Visit Come Together Media LLC to review the Sprint, Office Hours, and the tracking-first approach I use to rebuild healthcare campaigns.