How medical practices in the US set up search ads that book

Tuesday at 7:40am in Phoenix, a parent types “pediatric urgent care open now” while their kid runs a fever. In Houston, a new patient searches “dermatologist accepting new patients near me” between meetings. In Chicago, someone who just got an insurance card Googles “primary care doctor same week appointment.” Those are not awareness moments. They are booking moments, and they almost always start on Google Search.

For US doctors practices (primary care, specialty clinics, multi-provider groups), paid search is still one of the few channels where intent is loud enough to hear. People are not scrolling for inspiration. They are choosing who gets the call, the portal request, or the form that becomes a scheduled visit. That is what paid search for medical practices has to optimize for: cost per booked appointment, not a pretty CPL that dies in the front desk queue.

This is a how-to for owners and marketing leads who want Google Ads to fill the schedule without lighting money on broad terms, weak landing pages, and leads nobody calls back. You will map high-intent queries, structure campaigns the way a clinic actually works, protect spend with negatives, decide call vs form, match the landing page to the ad, and measure booked work instead of vanity form fills.

Why clinic search ads fill the phone and still fail the schedule

Most doctor accounts do not fail because “Google is too expensive.” They fail because the metric and the operation are misaligned. You bid on “doctor near me,” collect $45-$120 form leads (directional; specialty and metro vary hard), then watch no-shows, wrong insurance, wrong service line, or a three-hour callback lag kill the visit. The ad platform still calls it a conversion. Your calendar does not.

High-intent searches that actually produce booked work look specific: service + urgency or qualification. Think “knee replacement consult Dallas,” “allergist new patient appointment,” “telehealth psychiatrist accepting Aetna,” “walk-in clinic open Sunday,” “cataract surgeon near me.” Compare that to “best doctor” or single-word specialty terms that pull research, students, and job seekers. If 20-30% of SEM budgets routinely leak to irrelevant queries and weak landing match (a pattern we still see across verticals), clinics feel it as front-desk chaos, not as a tidy wasted-spend chart.

Three failure modes show up again and again in US practices:

• Wrong intent in the account. Campaigns mix informational and transactional queries. Smart bidding optimizes for whatever you labeled a conversion, including newsletter signups or “contact us” spam.

• Call vs form mismatch. Mobile drives a large share of paid search clicks, but if the ad pushes a long form for “urgent care open now,” you lose the patient who would have dialed in 20 seconds.

• Response capacity ignored. Dayparting is off, or ads run when the phones roll to voicemail. A lead at 9:15pm with a 14-hour callback is not a lead. It is a competitor’s patient.

Landing pages make it worse when ads promise “same-week new patient appointments” and the page is a generic homepage with a buried phone number, no insurance language, and a 12-field form. If someone on your team says they “just send traffic to the homepage,” treat that as a red flag. Dedicated pages aligned to the ad claim are table stakes for SEM / Google Ads in healthcare local markets.

When post-click proof is fuzzy, watch real sessions. A short pass with HeyLead Insights (session recordings, heatmaps, scroll and form behavior) usually shows the exact leak: patients stall on insurance copy, rage-tap a non-clickable phone, or abandon after the third required field.

If you want a partner already fluent in clinic lead gen rather than generic lead volume, skim how we approach Doctors marketing and keep reading for the build sequence you can run yourself first.

Map intent before you open Google Ads: queries that book visits

Do not start in the campaign UI. Start with a one-page intent map your medical director and front desk would recognize.

DIY playbook: intent map for a US doctors practice

• 1List every bookable “job”: new patient primary care, specialist consult, procedure eval, urgent/same-day, telehealth, workers’ comp or occupational if you offer it, self-pay cash packages if relevant.

• 2For each job, write 8-15 search phrases a real patient would type, including city or neighborhood modifiers you actually serve (e.g. Scottsdale, Buckhead, Naperville), insurance or “accepting new patients,” and urgency (“today,” “this week,” “walk-in”).

• 3Tag each phrase: Book-now, Research, Wrong-service, Careers/education, Competitor-brand. Only Book-now gets primary bid attention in week one.

• 4Pull Search Terms from any legacy account or use Keyword Planner as a sanity check. Kill ego terms with high volume and zero schedule impact.

• 5Define one primary conversion: booked appointment (CRM or scheduling status), with call conversions secondary if the call center or front desk can confirm disposition.

• 6Set a working ceiling for cost per booked visit from your contribution margin, not from a blog “average CPL.” If a new patient LTV supports $180 to acquire in your specialty, say that out loud before you chase a $40 form fill.

DIY tools for this section

• SEO competitor analysis (landing gaps)

• UTM link builder

• SERP snippet preview

• Meta description checker

US English, US metros, USD: write ad copy the way patients talk. “Accepting new patients,” “most insurance plans,” “same-week appointments,” “board-certified,” “on-site imaging” beat vague “comprehensive care” every time. Keep HIPAA and claim discipline in mind: no guaranteed outcomes, no fabricated wait times, no implying a relationship with a payer you do not have.

Build the Search account so front desk capacity and ads move together

Structure follows service lines and intent, not a pile of tightly themed ad groups from 2019 that you never negative out. Fewer campaigns, clean conversion setup, and honest bidding beat constant reshuffles that reset learning.

• 1Campaign split by bookable line Separate Search campaigns for high-value lines (e.g. ortho consults vs primary care new patients vs urgent care). Shared budgets across unrelated lines hide which specialty is actually profitable. Use location targeting to your real catchment; exclude ZIPs you will not serve.

Separate Search campaigns for high-value lines (e.g. ortho consults vs primary care new patients vs urgent care). Shared budgets across unrelated lines hide which specialty is actually profitable. Use location targeting to your real catchment; exclude ZIPs you will not serve.

• 2Match types with room to learn Start with exact and phrase on your Book-now list. Add carefully observed broad only after negatives and conversion tracking are trustworthy. Google still needs keywords and negatives even as bidding gets more signal-driven. “Signals over keywords” is not “delete the keyword list and pray.”

Start with exact and phrase on your Book-now list. Add carefully observed broad only after negatives and conversion tracking are trustworthy. Google still needs keywords and negatives even as bidding gets more signal-driven. “Signals over keywords” is not “delete the keyword list and pray.”

• 3Negative keyword discipline weekly Seed negatives on day one: jobs, salary, residency, “how to become,” DIY, veterinary if you are human medicine, free clinic if you are not, unrelated body parts, competitor hiring terms. Review search terms at least weekly for the first 45 days. Ask any agency “how do you handle negative keywords?” If the answer is vague, your budget will educate strangers.

Seed negatives on day one: jobs, salary, residency, “how to become,” DIY, veterinary if you are human medicine, free clinic if you are not, unrelated body parts, competitor hiring terms. Review search terms at least weekly for the first 45 days. Ask any agency “how do you handle negative keywords?” If the answer is vague, your budget will educate strangers.

• 4Call vs form by query type Urgent and same-day queries: call extensions, click-to-call, short mobile landing pages with the number sticky. Elective consults and insurance-heavy new patients: form plus phone, with fields that match scheduling reality (insurance, preferred location, reason for visit) without a novel-length intake. Track calls with a dedicated forwarding number and call length thresholds that mean a real conversation, not a 6-second hangup.

Urgent and same-day queries: call extensions, click-to-call, short mobile landing pages with the number sticky. Elective consults and insurance-heavy new patients: form plus phone, with fields that match scheduling reality (insurance, preferred location, reason for visit) without a novel-length intake. Track calls with a dedicated forwarding number and call length thresholds that mean a real conversation, not a 6-second hangup.

• 5Dayparting tied to response capacity If phones are covered 8am-5pm weekdays plus Saturday morning, do not bid like a 24/7 call center. Raise bids in peak booking windows; cut or pause when voicemail owns the line unless you have true after-hours clinical triage. Align ad schedules with the same spreadsheet ops uses for staffing.

If phones are covered 8am-5pm weekdays plus Saturday morning, do not bid like a 24/7 call center. Raise bids in peak booking windows; cut or pause when voicemail owns the line unless you have true after-hours clinical triage. Align ad schedules with the same spreadsheet ops uses for staffing.

• 6Bidding without fantasy ROAS promises New or thin accounts often need manual or limited automated bidding until conversion volume is real. Automated bidding can inflate spend when the conversion is a low-quality form. Industry ROAS benchmarks vary widely by category - healthcare is not retail. Optimize to cost per booked appointment and show rate, then let bidding automate on clean data. Nobody can honestly guarantee a CPL before seeing your account, pages, and unit economics.

New or thin accounts often need manual or limited automated bidding until conversion volume is real. Automated bidding can inflate spend when the conversion is a low-quality form. Industry ROAS benchmarks vary widely by category - healthcare is not retail. Optimize to cost per booked appointment and show rate, then let bidding automate on clean data. Nobody can honestly guarantee a CPL before seeing your account, pages, and unit economics.

• 7RSA copy that mirrors the landing claim Headlines: specialty, “accepting new patients,” city, insurance or self-pay clarity, appointment speed you can keep. Sitelines to locations, patient portal, insurance list, providers. Preview snippets so you do not look like every other clinic in Atlanta or Miami SERPs.

Headlines: specialty, “accepting new patients,” city, insurance or self-pay clarity, appointment speed you can keep. Sitelines to locations, patient portal, insurance list, providers. Preview snippets so you do not look like every other clinic in Atlanta or Miami SERPs.

• 8Conversion hygiene GTM + primary conversion on thank-you or confirmed schedule event. Enhanced conversions where appropriate. Strip micro-conversions from the primary goal. UTM everything so your EHR-adjacent CRM or practice software can show channel, not only “web.” Expect some signal loss from privacy tools; first-party and clean events matter more than ever for smart bidding.

GTM + primary conversion on thank-you or confirmed schedule event. Enhanced conversions where appropriate. Strip micro-conversions from the primary goal. UTM everything so your EHR-adjacent CRM or practice software can show channel, not only “web.” Expect some signal loss from privacy tools; first-party and clean events matter more than ever for smart bidding.

Performance Max can support Search later for feed and local inventory style presence, and many accounts say “Performance Max drives the bulk now.” For doctors booking visits from high-intent queries, win Search fundamentals first. PMax without negatives, brand controls, and offline booking import becomes a expensive mystery box.

Soft check: when the structure is live, pair it with landing pages that load fast. Only about 55.9% of origins pass all three Core Web Vitals in recent CrUX data, and mobile lag quietly taxes paid traffic. Run your candidate URLs through speed and tag checks before you scale spend.

Free checks before you raise budgets

• Core Web Vitals checker

• UTM link builder

• H1 checker

• Title tag checker

Landing pages and response: where paid clicks become charted patients

Ad groups should land on pages that continue the sentence the ad started. “Same-week dermatology consults in North Dallas” should not open a brand story from 2014. Above the fold: who you are, what visit type, proof (board certs, years, patient volume stated carefully), insurance or cash-pay path, primary CTA (call and/or book), secondary CTA for existing patients so they do not clog new-patient forms.

Proof that moves medical buyers is concrete: named conditions you treat, provider photos with credentials, parking and location clarity, languages spoken, average time-to-appointment if accurate, and reviews that mention wait times and bedside manner. Provider headshots with credentials and a named condition list outperform generic imagery - patients are vetting the doctor, not the brand.

Forms: reason for visit, insurance carrier (not full member ID on first step if that tanks completion), preferred campus, callback window. Multi-step can help if step one is fast. Instant Forms in other channels are a different animal; for Google Search, a fast first-party page usually keeps messaging, HIPAA-aware hosting, and CRM fields under your control.

Response speed is part of media. Route paid leads to a queue that knows they are paid. Script the first 60 seconds: confirm service line, insurance, urgency, offer two appointment slots. Measure time-to-first-touch in minutes, not “we try to call same day.” Clinics that treat speed as clinical operations (not a marketing afterthought) see the same media budget book more visits.

When conversion rate stalls, do not only tweak RSAs. Watch behavior on the page. Heatmaps and session recordings show whether patients ever see the insurance block, whether the click-to-call is broken on iOS, or whether the form error state is invisible. That is the practical use of HeyLead Insights next to your GA4 events: evidence over opinions in the marketing meeting.

Two US clinic scenarios: what changed in the account and on the floor

These are composite patterns we see across US clinic accounts-not named case studies. Use them for pattern recognition.

Multi-location primary care, Phoenix metro. Marketing was celebrating a $52 CPL on “primary care doctor near me.” Front desk said half the forms were existing patients resetting passwords or people outside the catchment. The fix was not a bigger budget. They split new-patient Search from brand, added negatives for portal, login, jobs, and out-of-area cities, rebuilt one landing page per campus with sticky click-to-call, and imported only “new patient scheduled” from the scheduling tool as the primary conversion. Bid schedule matched nurse-line coverage. Within a few weeks the reported CPL rose slightly (often in the single-digit to low-double-digit percent range in accounts like this), but cost per kept new-patient appointment dropped because junk stops looking like success. The mechanism was conversion definition plus campus-level pages, not a miracle keyword.

Specialty ortho group, suburbs outside Chicago. They ran broad specialty terms and sent everything to the homepage. Surgeons wanted “knee replacement” consult volume; ads attracted general sports-med browsers. They carved a Search campaign for procedure-intent phrases, wrote RSAs that named the procedure and “consultation,” required a short form plus phone, and trained schedulers to ask injury timeline and imaging status on first contact so the wrong acuity did not burn surgeon slots. Dayparting cut late-night spend when only an answering service picked up. Booked consult rate per click typically improved once message match and after-hours waste were fixed-often a noticeable lift within a few weeks when intent separation is tied to clinical capacity. Again, one mechanism: intent separation tied to clinical capacity.

Use these as pattern recognition, not as promised benchmarks. Your CPCs in Los Angeles will not match a secondary metro, and surgical consult value is not a sick-visit value.

What marketing leaders are seeing

“HeyLead helped us build, test and manage our paid SEM program. They were very responsive to our specific goals and brought subject matter expertise, saving us a lot of time and effort. They take a very practical results-oriented approach.” - Joe Charlson, CEO, CallerReady

FAQs

What metric should a medical practice actually optimize for in Google Ads?

Optimize for cost per booked appointment (and show rate if you can join the data), not raw lead volume. A cheap form fill that is wrong insurance, wrong campus, or never contacted is expensive. Feed scheduling outcomes back into Google as offline conversions when your stack allows it.

Should we use call-only ads or landing pages?

Use both by intent. Urgent and “open now” queries lean call-only or call-forward heavy. Elective and multi-location choices need a page that answers insurance, providers, and directions, with phone still obvious on mobile. Test; do not pick one format for the whole account.

How fast should staff respond to paid search leads?

Treat paid leads like a ringing phone in the lobby. Many clinics aim for minutes during open hours, not end-of-day batch callbacks. If you cannot staff that, shorten ad schedules. Paying for intent you cannot answer is how CPC inflation feels worse than it is.

Is Performance Max enough without Search campaigns?

Usually not for high-intent doctor booking. PMax can complement once tracking and creative assets are solid. Core service-line Search with negatives and tight landing match should remain the spine for queries that already mean “I need a doctor.”

What budget mistakes should we avoid early?

Do not promise leadership a fixed ROAS from week one. Do not let homepage traffic count as success. Do not skip negatives. Do not change structure daily. Give clean conversion data a few weeks, review search terms religiously, and scale what books kept visits.

Putting it to work

This week

• Write the intent map for your top three bookable visit types and mark Book-now phrases only.

• Export search terms (or draft the seed list) and build a day-one negative list with front desk input.

• Confirm one primary conversion = new patient scheduled (or equivalent), not generic form submit.

• Pick one service line and ship a dedicated landing page with sticky mobile call and honest insurance copy.

• Align ad schedule to real phone coverage; pause hours that only hit voicemail.

• Run your page through the Core Web Vitals checker and fix the worst mobile issue before raising bids.

Next 30 days

• Split campaigns by service line; stop one blended “all doctors” Search campaign if you still have it.

• Import or manually reconcile booked jobs weekly; kill keywords that never become charts.

• A/B the form length vs call emphasis on your highest-spend ad group.

• Document response SLAs and measure median first-touch time on paid leads.

• Only then test broader match or PMax with guardrails.

If you would rather not carry keyword hygiene, call tracking, landing proof, and booking-quality feedback loops alone, HeyLead can own the search-to-scheduled-visit path for US doctors practices (account structure, negatives, pages, and measurement tied to booked jobs). Start a conversation at [email protected].