A new patient from Google Ads is worth paying for only up to a ceiling, and most practices never calculate it. That ceiling is what a new patient contributes after payer mix, margin and capacity, divided by the return you want.

Work backward from it to the most you can pay per lead and per click, then check how many leads actually become patients. This page shows how, with a worked example, an input worksheet, and the places where the money usually leaks.

It is written for owners, partners and administrators of private practices and small groups across dental, medical, chiropractic, physical therapy, mental health and other specialties. It does not cover insurer, pharma or health-system advertising, which work differently.

What you want to knowFormula
Most you can pay for a new patientContribution value per new patient ÷ target multiple
Most you can pay for a leadAllowable cost per new patient × share of leads who become new patients
Most you can pay for a clickMax cost per lead × click-to-lead rate

How healthcare paid search works

Healthcare paid search means bidding on the searches patients type into Google, and sometimes Microsoft’s Bing, so your practice appears in the ads above or beside the results. A patient searches, sees your ad, lands on a page, and calls or fills in a form. Your front desk answers, books and checks coverage, and the patient comes in.

Money is made or lost at each step. The auction sets what the click costs. The page and the phone decide whether the click becomes a booking. Your schedule and payer mix decide whether the booking is worth having. Most reporting inside the ad platform covers only the first step or two.

What a new patient is worth, and what you can afford to pay for one

A good cost per new patient is one that leaves a healthy return after payer mix, margin and capacity. A rule of thumb borrowed from subscription businesses is to spend no more than about a third of a customer’s value on acquiring them. Treat it as a starting point and adjust.

Step 1: Estimate contribution value per new patient

Contribution value is what a new patient adds after the direct costs of caring for them, before the cost of acquiring them.

  • Use collected revenue, not billed charges. Collections reflect what payers and patients actually pay.
  • Pick a time window and stay consistent. Twelve months is a reasonable default. Include the follow-up care a first visit typically leads to, but do not assume lifetime loyalty.
  • Subtract variable costs of delivering that care, such as supplies and lab fees. Rent and other fixed overhead stay out, because they exist whether or not this patient arrives.
  • If clinical labor is salaried, or you are the clinician, treat it as a fixed cost when you have open capacity and as an opportunity cost when you don’t. A new patient who takes a slot that would have gone to a more valuable visit is worth less than the revenue suggests.
  • To find the number, pull a cohort of patients whose first visit was at least twelve months ago from your practice management system, total their collections by payer, and divide by the number of patients. If you haven’t tracked where patients came from, use all new patients as a stand-in and label it that way.

Then split by payer, because the same service pays differently depending on who is paying. Use the payer mix of the patients paid search actually brings in, which may differ from your overall mix.

Hypothetical example (illustrative inputs, not benchmarks):

Payer groupShare of new patients12-month collected revenue per patientWeighted
Commercial insurance50%$2,400$1,200
Medicare30%$1,500$450
Self-pay20%$1,750$350
Weighted collected revenue$2,000

Assume a 60% contribution margin. Contribution value per new patient is $2,000 × 0.60 = $1,200.

If you are out-of-network with a large payer, model that payer at your actual collections, or leave it out of your advertising altogether. The section on which services to advertise returns to this.

Step 2: Choose a target multiple

Decide how many dollars of contribution value you want back for each acquisition dollar. A multiple of 3 means you spend about a third of a new patient’s contribution on acquiring them and keep the rest for fixed overhead, profit and the chance your estimates are off. A practice with a long payback tolerance might accept less. A practice with tight capacity might demand more.

Mind the timing, too. You pay the acquisition cost up front and collect the contribution over the year. A practice with thin cash reserves may need a shorter window than twelve months.

At a multiple of 3, the example practice can afford $400 per new patient ($1,200 ÷ 3).

Step 3: Work backward to leads and clicks

You need four rates from your own data:

  • Click to lead: the share of ad clicks that become a call or form
  • Valid: the share of leads that are real, local new-patient inquiries for a service you offer
  • Booked: the share of valid leads that get scheduled
  • Show: the share of booked patients who attend

The share of leads who become new patients is valid × booked × show.

Hypothetical example, continued. Say 70% of leads are valid, 55% of valid leads book, and 80% of those show. That is 0.70 × 0.55 × 0.80 = 30.8% of leads becoming new patients.

  • Max cost per lead: $400 × 0.308 = $123
  • With a 7% click-to-lead rate, max cost per click: $123 × 0.07 = $8.62

Now suppose better call handling raises the booked rate from 55% to 75%. Lead-to-patient becomes 0.70 × 0.75 × 0.80 = 42%.

  • Max cost per lead: $400 × 0.42 = $168
  • Max cost per click: $168 × 0.07 = $11.76

Same ads, same landing page, same service. The practice can afford to bid about 36% more per click because its front desk converts more of the leads it receives.

Where to find each input

InputWhere it comes from
Weighted collected revenue per new patientPractice management system cohort, by payer
Contribution marginRevenue minus variable cost of care, divided by revenue
Target multipleYour decision. Start at 3
Valid-lead rateGrade 50 recent leads (see the funnel section)
Booked rateBooked appointments ÷ valid leads
Show rateAttended ÷ booked
Click-to-lead rateLeads ÷ ad clicks

If you don’t have a number yet, estimate it, label it an estimate, and use the 30-day plan at the end to replace it with a measured figure.

Why healthcare PPC benchmarks disagree, and what to use instead

Published figures conflict because they come from different report editions, categories and campaign samples, and because a “lead” in an ad account is a setting you chose, not a patient. Use benchmarks as a sanity check, not a target.

Most of the numbers that circulate trace to WordStream by LocaliQ. Here is what its own reports show for healthcare-related categories.

ReportCategoryMedian CPCConversion rateCost per lead
2025 editionDentists & Dental Services$7.859.08%$83.93
2025 editionPhysicians & Surgeons$5.0011.62%$56.83
2025 editionHealth & Fitness$5.006.80%$62.80
2025 editionAll industries$5.267.52%$70.11
2026 editionDentists & Dental Servicesnot on that page10.67%$72.97
2026 editionPhysicians & Surgeonsnot on that page12.43%$40.04
2026 editionHealth & Fitnessnot on that page6.94%$67.36
2026 editionAll industriesnot on that page8.18%$66.69

The 2025 report analyzed 16,446 US campaigns from April 2024 through March 2025 across Google and Microsoft search ads, with at least 64 campaigns per category. It describes its averages as medians. The sample behind the 2026 figures was not shown on the page reviewed.

Four things in this table explain most of the confusion:

  • Edition. Between the two reports, the median cost per lead fell by roughly 30% for physicians and 13% for dentists. Comparing a 2024 number with a 2026 number mixes different years.
  • Category. “Physicians & Surgeons,” “Dentists & Dental Services” and “Health & Fitness” are different auctions. Physicians show a conversion rate of 12.43%, while Health & Fitness shows 6.94%.
  • Platform and sample. The 2025 data blends Google and Microsoft campaigns, and each edition has its own sample.
  • What a conversion is. WordStream treats cost per lead as cost per conversion and notes that it depends on how advertisers track and value their conversion actions. A tap on a call button and a booked appointment can both be counted as a “lead.”

Other numbers circulate that don’t trace back. Some vendor posts cite a “Health and Medical” category at $2.62 per click and a 3.36% conversion rate. That category does not appear in WordStream’s 2025 tables, so treat those figures as untraceable until someone produces the source. One vendor post gives $2.69 as the 2025 all-industry average click cost, while WordStream’s own report gives $5.26.

Turning a cost per lead into a cost per patient

A cost-per-lead benchmark says nothing about cost per new patient until you apply your own funnel. Take WordStream’s 2026 median cost per lead and run it through two hypothetical funnels:

Category (2026 median CPL)If 30.8% of leads become new patientsIf 15% do
Dentists & Dental Services, $72.97about $237about $486
Physicians & Surgeons, $40.04about $130about $267

The same published cost per lead produces cost per new patient figures that differ by roughly a factor of two, depending on what happens after the lead arrives. These are illustrations, not benchmarks.

The research behind this page turned up no primary-source benchmark for cost per new patient, and the figures that circulate for it are mostly agency claims with undefined terms.

What to use instead. Build your own baseline from the last 90 days of calls and forms (the 30-day plan shows how). Compare your cost per lead to the median for your category. Then calculate your own lead-to-patient rate. Before you rely on any benchmark, ask four things: which edition, which category, which platforms, and what counted as a conversion.

Where the patient funnel leaks after the click

Cost per new patient equals cost per lead divided by the share of leads who become patients. A cheap lead that rarely becomes a patient is not cheap.

Here is the same hypothetical practice, spending $6,000 a month at a $9 CPC with the rates above (7% click to lead, 70% valid, 55% booked, 80% show). The right-hand column changes one thing: the booked rate rises to 75%.

StageBaselineBooked rate at 75%
Spend$6,000$6,000
Clicks ($9 each)667667
Leads (7%)4747
Valid new-patient leads (70%)3333
Booked1824.5
Showed (80%)14.419.6
Cost per new patientabout $417about $306
Paired funnel chart for a hypothetical practice: 667 clicks become 14.4 new patients at baseline and 19.6 after the booked rate rises to 75%, cutting cost per new patient from about $417 to about $306.

Compare the baseline with the $400 ceiling from the first section. At a $9 click cost, this practice pays $17 more than it can afford for every new patient. Fixing handling brings it $94 under the ceiling, a reduction of roughly 27%, and nobody touched a bid. These inputs are illustrative. Yours will differ.

The same arithmetic explains a common trap. A lead that costs $100 and books half the time is a $200 patient before no-shows.

Diagnose from the right

Work backward from revenue. The rightmost broken stage multiplies everything to its left, so fix it first.

SymptomLikely causeFirst check
Good cost per lead, schedule still openConversion event is too shallow, or leads are unqualified or unansweredGrade the last 50 leads by outcome (below)
Strong reported results, new-patient count flatBrand searches and existing patients inflating conversionsSplit brand from non-brand and classify callers as new or existing
High click-through, low conversionThe ad promises something the page doesn’t deliverRead the ad, then the page headline and booking path, as a patient would on a phone
Plenty of valid leads, few bookedSlow or missed response, scripting, insurance confusionMeasure time to first contact and live-answer rate, including after hours
Booked, but many no-showsLong wait to appointment, no reminders, coverage surprisesShow rate by days between booking and appointment
Patients show, revenue per patient is lowPayer or service mix, out-of-network patientsRevenue per new patient by campaign and payer
Automated bidding drifts toward the wrong trafficThe signal it learns from is weak or noisyAudit which conversion actions it optimizes toward

How to grade 50 leads. Pull the last 50 calls and form submissions. Sort each into one of four groups: new patient, good fit, booked; new patient, good fit, not booked; existing patient or referral source; not a fit (wrong service, wrong area, spam, vendor). The proportions give you your valid-lead and booked rates, and the “not booked” group shows you where to look first.

How much should a medical practice spend on Google Ads?

Work backward from how many new patients you can see each month. Divide by the share of leads that become patients, then by your click-to-lead rate, then multiply by your cost per click. Check the result against local search volume and your capacity.

Using the same hypothetical practice, here is the budget for 20 new patients a month:

Calculation stepBaselineAfter handling fix
Leads needed (20 ÷ lead-to-patient rate)about 65about 48
Clicks needed (leads ÷ 7%)about 928about 680
Monthly budget (clicks × $9)about $8,350about $6,120

The same 20 patients cost about $2,200 less each month once the front desk converts more of the leads it already gets. That gap is the reason to learn your rates before you set a budget.

Two checks before you commit:

  • Is there enough demand? Paid search captures searches that already exist. In a small market, or for a rare service, there may not be enough qualified searches to spend this much at an acceptable cost.
  • Can you see these patients? If your schedule is full, the budget buys waiting lists and cancellations, not revenue.

Which searches to buy

Buy searches in the order that matches your economics, not your keyword tool’s volume numbers. Different kinds of searches carry different intent, and that changes cost per patient more than it changes cost per click.

Search typeExampleRoleEconomic implication
Symptom or question“why does my knee hurt”Early researchBroad and low commercial intent. Mostly waste unless a service-specific page fits
Treatment or procedure“physical therapy after ACL surgery”Informed researcherNarrower and more relevant. Good for specific service lines
Provider, near me, accepting new patients“orthodontist near me accepting new patients”Ready to chooseThe core of most direct-to-patient campaigns. Clicks cost more, but more of them become patients
Insurance-qualified“dentist that takes [plan]”Payer fit built into the queryStrong when you accept the plan. Wasted when you don’t
Urgent“emergency dentist open now”Needs care todayRequires call coverage and an after-hours plan
BrandYour practice or a physician’s nameExisting demandReport separately. It rarely adds new patients

The principle behind the table is reasoned, not measured: higher-intent searches cost more per click and tend to cost less per booked patient, because more of the people who click arrive ready to book. Your own lead grading will confirm or contradict that for your market.

Three practical habits keep spend clean:

  • Read the search terms report weekly at first. It lists the actual searches that triggered your ads. Broad match can pull in the wrong services, areas and payers, so add negatives as you find them. Common starters are job-seeker terms (“jobs,” “salary,” “training”), services you don’t offer, and payers you don’t accept.
  • Look at your own results. Search your top terms from a signed-out browser. Directories, listing sites and insurer provider finders often appear beside your ads and compete for the same clicks, which affects what you can expect to pay.
  • Treat language as a targeting decision. If a meaningful share of your patients searches in Spanish, check your search terms and call data first, then build Spanish campaigns and pages instead of translating only the ad.

People who click an ad often check your Google Business Profile and organic listing before they call, which is why local SEO and paid search tend to be judged together by patients, even when they are budgeted separately.

Which services should run paid search

Advertise a service when it clears five gates. A campaign that fails a gate can look good in the ad platform and still lose money.

  1. Margin. After payer mix and delivery cost, is contribution value high enough to support a meaningful acquisition cost? If not, fix pricing, mix or scope first.
  2. Capacity. Can you see these patients within a reasonable time without displacing more valuable visits? If the schedule is full, throttle or pause.
  3. Payer fit. Are you in-network for the plans most local searchers carry? If not, exclude the searches that attract those plans and avoid copy that implies otherwise.
  4. Search volume. Is there enough local demand to support the spend?
  5. Policy and referral dependency. Is the category restricted by Google’s policies, or do most patients arrive through a referring physician?
Flowchart of five gates a service must clear before running paid search: margin, capacity, payer fit, search volume, and policy or referral dependency, with a fix or restriction for each failure.

Insurance and payers in the funnel

In the United States, insurance acts as an early filter in how patients choose, so it belongs in the funnel, not only the billing office.

  • Put accepted plans on the landing page, with a line telling visitors to verify coverage.
  • Name Medicare Advantage plans specifically. Each has its own network, so “we accept Medicare” can leave patients guessing.
  • Do not write “we accept all insurance” unless it is true.
  • Verify coverage during intake, before the visit, to protect your show rate and collections.
  • Offers to Medicare or Medicaid patients, such as free services or waived copays, raise federal and state questions. Ask counsel before using them.

Restricted categories

Mental health, addiction treatment and some cosmetic and fertility services face tighter rules on targeting, claims and certification. Google requires certification for addiction treatment advertising in the United States, and rules change, so check Google’s current policy before you launch. If you offer telehealth, target only the states where the provider is licensed, since licensure rules vary by state.

Referral-gated specialties

Where care usually starts with a referral, as in many oncology, cardiology, orthopedic and GI services, general acquisition spend may reach only part of the decision. A plausible hypothesis, not a measured finding: paid search suits narrower roles here, such as procedure-specific research, second opinions and protecting your own name in search results.

Test small and measure before assuming a playbook built for direct-to-patient services applies. Specialty-specific marketing differs enough that the budget logic should be run per service line, not per practice.

What to tell the bidder: choosing your conversion

Google Ads gets better at finding more of whatever you tell it counts as a conversion. Choose the deepest event you have enough volume to optimize toward, and report on something deeper still. Call this the conversion ladder.

EventWhat the bidder learnsRisk
Click on a phone numberSomeone intended to callCounts dials that never connect
Call lasting over a set lengthA conversation happenedIncludes existing patients and wrong numbers
Form submissionSomeone left detailsIncludes spam, vendors and tire-kickers
Qualified new-patient inquiryA real prospect for a service you offerRequires someone to classify it
Booked appointmentA patient got on the scheduleFewer events, so slower learning
Showed or paidA real patient arrivedRare events, a delay, and system integration

Optimize on one event, report on another. In Google Ads you can mark each conversion action as primary or secondary. Primary actions feed bidding. Secondary actions are tracked and reported without steering the bidder. That lets you optimize toward a call over a minimum length while still watching booked appointments.

If volume is thin. Many practices generate only a handful of bookings a month. Automated strategies such as Target CPA or Maximize Conversions need enough conversion data to learn from, and you will often see a rule of thumb of 30 to 50 conversions a month.

It is a rule of thumb, not a documented requirement, so check Google’s current guidance. When volume is too low, optimize toward a shallower event that tends to precede bookings, and track bookings and shows in your own scheduling system.

Different services, different values. Value-based bidding lets you tell Google that one service line is worth more than another. It needs reliable value data and enough volume, so it is usually a later step.

Calls. Call tracking swaps the phone number a visitor sees depending on where they came from, which is how calls get tied to campaigns. Call recording is covered by state consent laws that differ from state to state. Ask any call-tracking vendor whether it will sign a business associate agreement.

Bringing outcomes back. Offline conversion import tells Google which clicks became booked patients. One design to discuss with counsel is to store the ad’s click ID with the lead in your own system, then upload only that ID, the conversion time and a generic “booked” label, with no name, condition or service.

Whether even that is acceptable for your practice is a legal question. Enhanced conversions work differently and send hashed personal identifiers, which are still personal data.

Brand search and existing patients. Searches for your practice or a physician’s name often come from people who already chose you, including current patients and referrals. Those campaigns tend to look excellent on cost per conversion and add the least new demand.

Keep brand and non-brand in separate campaigns and judge growth on the non-brand side. Brand coverage still has defensive value if competitors bid on your name, and you can compare it against how you appear in organic listings for the same terms.

Landing pages and the booking path

A landing page for a healthcare ad has one job: make booking the obvious next step on a phone without collecting more than it needs.

  • Message match. The headline repeats the promise in the ad.
  • Click-to-call at the top, with hours and a clear path when the office is closed.
  • An accurate “accepting new patients” status.
  • Location and service area, with a map or ZIP code line.
  • Accepted insurance, with a “verify coverage” line.
  • Provider credentials, presented as your state board permits.
  • Speed on mobile. Test it on a cellular connection, not office Wi-Fi.
  • A short form. Ask for name, phone and preferred time. Avoid free-text fields that invite visitors to describe symptoms unless you have reviewed what happens to that text.
  • A neutral confirmation page and URL. A thank-you page named after a condition tells every tool that reads the page why the visitor came.
  • Honest proof. If you display reviews or testimonials, the FTC’s rule on consumer reviews and testimonials has applied since October 2024, and your state board may add limits. Ask all patients for reviews the same way, and don’t promise outcomes.

Many practices handle this through healthcare website design that treats booking pages as part of the ad system, not as brochure pages.

Tracking, privacy and ad policy

Review note. This section describes the landscape and is not legal advice. Following it does not make a setup compliant. Have qualified healthcare privacy counsel review what your site, forms, call tracking and ad accounts actually transmit.

StatusPointWhy it matters
EstablishedHIPAA applies to covered entities (most health plans, clearinghouses and providers who bill electronically) and their business associatesIf you are a covered entity, protected health information (PHI) is regulated wherever it goes
EstablishedGoogle’s business associate agreement (BAA) applies only to the products named in its HIPAA documentationSee below
Platform policyGoogle limits personalized advertising based on health conditions and requires certification for addiction treatment ads in the USAudience tools are restricted. Platform rules change
ContestedThe scope of HHS’s tracking-technology guidance after a 2024 federal court ruling, and how private wiretap and privacy suits apply to ad tagsSee below
EvolvingState health-privacy laws, including Washington’s My Health My Data Act and Nevada’s SB 370Some reach beyond providers and require consent. Washington’s law has carve-outs for data already covered by HIPAA, so confirm scope with counsel

Is Google Ads HIPAA compliant?

Google’s BAA covers the Google Cloud and Workspace products listed in its HIPAA documentation, and the agreement itself says it does not apply to any product not on the list. Advertising products did not appear in the excerpts reviewed for this page, and several third-party sources say Google Ads, Google Analytics and Google Tag Manager are not covered. Confirm against the full current list.

The practical consequence is that you should not send PHI to those tools. Whether your setup does depends on what your site and forms collect and whether you are a covered entity. That is a question for counsel, not a feature of the ad platform.

The law is moving. In June 2024, a federal court in Texas vacated part of HHS’s guidance on online tracking technologies, as summarized by a healthcare law firm. The ruling addressed the theory that tracking on unauthenticated public pages is covered by itself. Law-firm summaries say other parts of the guidance remain, including trackers on pages where patients log in.

How the ruling applies to pages where visitors enter appointment or condition details is a counsel question. Most of the best-known lawsuits involve Meta Pixel on hospital sites and patient portals, so applying them to other tags is an extrapolation, though outcomes in this area have been mixed.

Can practices retarget website visitors?

Google’s policy limits personalized advertising based on health conditions, and sources describe remarketing and similar audiences built from health-related page visits as restricted. Even where a tool is technically available, HIPAA and state privacy laws may apply. Clear it with counsel and check Google’s current policy before you build any list.

Send the event, not the context

The ad platform needs to know that a valuable outcome occurred. In principle, it does not need to know why the patient came in. Two habits follow:

  • Watch your URLs and event names. A page at /anxiety-treatment/thank-you or an event called booked_depression_consult can reveal a condition to every tool that reads it. Neutral naming and generic confirmation pages reduce what leaves your site.
  • Collect only what the conversion needs. Every extra form field and parameter is data that may end up somewhere you did not intend.

Server-side tagging is often presented as a compliance fix. It changes where data is processed, not whether sensitive data is shared, so treat it as a tool, not a legal solution.

Where paid search fits among your options

Build campaigns around services and their economics, start with Search, and treat newer automated formats as tests with guardrails.

OptionWhat it doesWatch for
Search adsCaptures people already looking for a service near themQuery drift: broad matching can pull in wrong services, areas and payers
Local Services AdsPer-lead pricing and a Google screening or guarantee badge where your category qualifiesEligibility varies by specialty, so verify it. Grade the leads like any other
Performance Max and AI MaxReach across Google surfaces driven by automationWeak signals produce wasteful learning, and creative must still meet health-ad policy. Run as a controlled test
Microsoft AdvertisingA smaller search audience that WordStream says typically costs less than GoogleLower volume. Check health-ad policy and test with a small budget
Paid socialReaches people before they searchWordStream’s 2026 medians show a lower cost per lead on Facebook lead campaigns for physicians ($32.14 vs. $40.04) and dentists ($61.56 vs. $72.97), but those are form leads, and the report notes search leads often carry stronger intent. Pixel-related privacy risk applies
Local SEO and Google Business ProfileBuilds visibility that compounds without a per-click costSlower to start and harder to attribute

A few structural habits follow from the economics:

  • Separate campaigns by service line, so each can be judged against its own allowable cost.
  • Match the ad schedule to call coverage. Paying for calls that go to voicemail is paying to lose them.
  • Keep geography realistic. Target by ZIP code or radius to the areas patients actually travel from, and exclude the rest.
  • Use location and call assets, so ads show your address and a tap-to-call option, and keep your Business Profile consistent with them. If you want help setting up managed Search and Local Services Ads campaigns, that is the kind of structure a specialist builds.

Multi-location groups. Overlapping geographies split data and budget across campaigns competing for the same searches. Give each location its own capacity numbers, decide whether intake is central or local, and make sure each segment generates enough conversions to learn from.

An emerging factor. A trade publication reported in June 2026 that Google began a limited test of healthcare ads in AI Mode in the United States. How it will affect practices is unknown, and the status may have changed.

When to scale, hold or stop

The average cost of a new patient can look fine while the last dollar you spent bought patients at a loss.

Paid search draws on a finite pool of local searches. The best searches are exhausted first, so each added dollar tends to reach a lower-intent audience or a more expensive auction position.

Hypothetical example (the practice from earlier after its handling fix, with its $400 ceiling, assuming diminishing returns as spend rises):

Monthly spendNew patientsAverage cost per patientCost of the added patients
$3,00011$273n/a
$6,00020about $300$333
$9,00024$375$750

The $6,000 row matches the funnel table above, rounded. At $9,000 the average ($375) is still under the $400 ceiling, but the extra $3,000 bought four patients at $750 each.

Three conditions to hold or stop:

  1. The cost of added patients exceeds your allowable cost, after you have fixed handling and tracking leaks.
  2. Impression share, the share of eligible searches where your ad actually showed, is already high for your best searches, so more budget has nowhere productive to go.
  3. Capacity is full. More leads will not become more seen patients.

How long before you can judge results? Long enough for conversions to mature. Urgent care books fast. Elective and higher-cost care takes longer, so recent weeks will always look worse than they end up. Judge cohorts after your typical lag.

When a campaign hits its ceiling, growth has to come from new services, new areas or other channels, such as organic social and local visibility, not from a bigger budget on the same searches.

Reading a report, and deciding who runs it

A report is useful if you can check it against your schedule and your bank account. It should show spend by campaign and service line, leads graded by quality, booked, showed and new patients, revenue per new patient, brand and non-brand separately, and the exact definition of each conversion action. Clicks, impressions, cost per lead with no definition of “lead,” and percentage improvements with no baseline flatter on their own.

Ask any provider, including an in-house hire:

  • Who owns the Google Ads account and the tracking? The practice should. Otherwise leaving means losing history.
  • How is a “new patient” defined, and does reporting go beyond cost per lead to booked and shown patients?
  • Can I hear recorded calls and see form submissions, within applicable recording-consent rules?
  • What is being sent to the ad platform, and who reviewed it?

The choice between in-house, specialist or both depends on the practice:

FactorLeans in-houseLeans toward specialist PPC management or a hybrid
Locations and service linesOne location, few servicesMultiple locations or many service lines
CategoryUnrestrictedRestricted category or heavy policy review
MeasurementCalls and forms onlyBooked and shown import, privacy review
TimeSomeone reviews leads weeklyNo one owns lead review
CapacityStableFrequently full or variable

Many practices land on a hybrid: a specialist for structure and measurement, and an internal owner for lead handling and the schedule.

Your first 30 days

The order matters. Fix the stage that multiplies the others first.

Week 1: Establish the numbers

  1. Calculate contribution value per new patient by payer group, using collected revenue.
  2. Set a target multiple and compute allowable cost per patient, lead and click.
  3. Pull 90 days of calls and forms from every source.

Week 2: Measure the funnel

  1. Grade 50 recent leads using the four categories above.
  2. Measure live-answer rate and time to first contact, including evenings and weekends.
  3. Calculate booked and show rates.

Week 3: Fix measurement and handling

  1. Define your conversion actions and separate new from existing patients.
  2. Separate brand from non-brand.
  3. Fix the biggest handling gap you found, whether that is call coverage, response speed or scripts.
  4. Have counsel review what your site sends to ad platforms before adding any outcome data.

Week 4: Tune what is left

  1. Check landing pages on a phone for message match and a clear booking path.
  2. Review search terms and add negatives for payers, services and areas you do not serve.
  3. Check each service line against the five gates.
  4. Only now adjust bids, budget and creative, using your allowable cost as the ceiling.

Where Pracxcel fits

If you would rather have someone check your numbers than build the model alone, that is a reasonable use of a specialist. According to its PPC service page, Pracxcel runs paid search campaigns, including Local Services Ads, with call, form and booking tracking. The agency organizes its work around the P.A.T.I.E.N.T. Framework, and several of its steps correspond to the ideas above:

  • Plan & Audit: unit economics and service-line gates
  • Attract: search structure and search intent
  • Integrate: landing pages, call tracking and the booking path
  • Nurture: speed and quality of follow-up with new inquiries
  • Track & Recall: reporting booked and shown patients against spend

The question this page asks of any provider applies here too: does reporting go beyond cost per lead to booked and shown patients, and how is a new patient defined? Bring your payer mix, your lead outcomes and your current spend to a PPC strategy call and ask exactly that.

Terms used on this page

TermMeaning here
New patientA first-time patient who is seen. Excludes existing patients, referral sources, vendors and wrong numbers
LeadAny call or form submission from an ad, before anyone checks who it is
Valid leadA real, local, new-patient inquiry for a service you offer
ConversionWhatever event you tell the ad platform to count. It is a setting, not a fact about patients
Cost per lead (CPL)Spend divided by leads
Cost per new patient (cost per acquisition)Spend divided by new patients who were seen
Contribution valueCollected revenue from a new patient over a stated window, minus the variable cost of delivering that care
Allowable costThe most you can pay for a new patient and still reach your target return
Impression shareThe share of eligible searches where your ad actually appeared

FAQs