Google Ads for clinics: why a cheaper lead can cost more

The short answer: a form submission is not the outcome

A cheaper Google Ads lead can cost your clinic more when it generates unsuitable administrative enquiries, repeated follow-up or bookings that do not attend. Compare campaigns using relevant enquiries and attended appointments, with a consistent allowance for the work needed to handle them.

Cost per lead still has a job: it tells you how much you paid for the event you defined. It does not tell you whether that event represents a new person, a useful request or an appointment. The definition beneath the number matters as much as the number itself.

This is an owner-and-manager measurement framework, not a promise of lower advertising costs. Google, ASA and ICO guidance was checked on 6 November 2026. Confirm current policies and your clinic's specific obligations before changing campaigns or tracking.

Agree the definitions before comparing agencies or campaigns

Define a raw enquiry as a submitted request, then separately count unique contactable enquiries. Remove staff tests, obvious spam and duplicates according to a documented rule. Do not quietly change that rule when a campaign's results look disappointing.

Administrative relevance might mean that the request concerns a service the clinic actually offers, the correct location and a contact route the person has supplied. It must not mean a marketer or receptionist has judged clinical suitability. That decision belongs in the appropriate professional process.

Keep booked, attended, cancelled, rescheduled and pending outcomes separate inside the clinic's authorised systems. A booking-link click is not a booking. An appointment booked yesterday cannot yet be fairly compared with one whose attendance date was last month.

A hypothetical comparison that changes the decision

Imagine two fictional campaigns evaluated after both enquiry groups have had time to reach their appointment dates. Each spends GBP 1,200 on ads. Campaign A produces twice as many raw enquiries, but also more irrelevant requests and more reception work. These figures illustrate arithmetic, not expected clinic performance.

Count the costs without pretending attendance equals profit

In the example, Campaign A's GBP 15 lead looks attractive until the clinic divides GBP 1,440 of advertising and handling costs by 12 attendances. Campaign B divides GBP 1,320 by 18. The campaign with the more expensive raw lead has the lower cost per attendance.

Use the same cost boundary for both campaigns. Include comparable agency, creative and landing-page costs where you can allocate them reasonably. Estimate reception effort from a small workload sample and an agreed staff cost, rather than assigning a convenient flat fee after seeing the result.

This is not a profit calculation. Clinical delivery costs, refunds, consultation fees and later decisions still matter. Do not assume every attendance becomes treatment, or build a campaign target around pressuring people into treatment. Commercial reporting must leave clinical judgement independent.

Read search intent before changing the bid

Review the available search terms with reception's recurring enquiry themes beside you. Separate requests to contact a local clinic from research, employment, training and supplier searches. A campaign can attract plausible-looking words while answering the wrong underlying question.

Check whether the ad and destination explain the real location, service and enquiry process. A visitor seeking an immediate appointment may be poorly served by a page that promises speed when the clinic has limited capacity. Fix the promise or the operational constraint before paying for more traffic.

Keep brand and non-brand performance visible separately. Someone already searching for your clinic may have heard about it elsewhere. Do not use a low-cost brand campaign to imply that all paid search is finding new demand at the same price.

Make a search-term decision sheet

For each available search term, record the apparent task, matching landing page and proposed action. 'Clinic receptionist jobs' may be irrelevant to a patient-enquiry campaign; 'consultation cost in [town]' may need clearer fee information, not exclusion. These are fictional examples, not recommended treatment keywords.

Add a reason and review date for each change. Also record what the search-terms report cannot show: it does not reveal every query, and campaign-level results may mix different intentions. Do not infer a person's diagnosis or clinical suitability from a search phrase.

Use negative keywords as a reviewed exclusion policy

Build exclusions from genuine mismatches, not a downloaded list applied without context. Employment and training terms may be sensible exclusions for a patient-enquiry campaign, but not for a campaign recruiting staff. Keep the reason and scope beside each important exclusion.

Google explains that negative keywords do not match close variants or other expansions in the same way as positive keywords. Review relevant singular, plural and synonym forms, and understand the difference between negative broad, phrase and exact matching before applying a change.

Avoid blocking useful enquiries through a broad term such as price simply because some price enquiries do not book. Clear pricing information may resolve uncertainty. Review search terms and reception feedback together, then make limited changes you can evaluate.

Qualify through clear information, not intrusive forms

State the branch, consultation arrangements, accurate fee context and how availability is confirmed before the enquiry button. This lets people decide whether the next step fits their practical needs without being interrogated by a marketing form.

Ask only for details required for the administrative response. Do not collect symptoms, medical histories or photographs to improve an advertising lead score. Give clinical questions an appropriate separate route. A shorter form is not automatically better if it hides essential information; a longer form is not automatically better at qualification.

Review the page on a phone. Check the telephone link, form errors, confirmation message and return route. A campaign cannot compensate for a form that fails or a page that sends everyone to the wrong branch.

Separate healthcare advertising rules from campaign economics

Google's healthcare and medicines policy restricts some content and applies certification and location-specific conditions to certain services. Do not assume that every clinic service is eligible because another advertiser appears in search results. Review the actual service, ad and destination against the current policy.

Google also restricts personalised advertising involving sensitive interests, including health. For advertisers promoting services in sensitive categories, advertiser-curated audiences are restricted. Do not turn patient lists or visits to sensitive service pages into remarketing audiences on the assumption that consent makes every targeting method permissible.

UK requirements apply separately. ASA guidance states that prescription-only medicines cannot be advertised to the public, including through sponsored search ads. Do not use an ad or linked landing page to promote a POM directly or indirectly. Obtain appropriate review rather than trying to bypass a restriction with different wording.

Keep clinical outcomes in the clinic, not in Google uploads

You can improve management reporting without uploading patient records to an advertising platform. Keep identifiable attendance and clinical information in approved clinic systems. Use suitably aggregated internal reports for the comparison, with access restrictions and care around small groups that could identify someone.

Do not send patient data, symptoms, diagnoses, treatment choices, appointment details or message contents to Google Ads or Analytics. Check page addresses, query strings, event names and form integrations as well as deliberate uploads. Hashing an identifier does not make a sensitive record anonymous or automatically eligible for advertising use.

The ICO says storage and access technologies used for online advertising require consent. Have the implementation reviewed before enabling advertising tags or associated tracking. A consent-management setting does not itself prove that the information collected is appropriate or that all supplier policies are satisfied.

Accept measurement gaps. Consent choices, telephone enquiries, repeat visits and different devices can limit attribution. An optional source question can provide context, but it is not perfect evidence either. Record unknowns rather than assigning every eventual appointment to the last visible ad click.

Investigate cancellations without blaming the channel

Compare like-for-like enquiry periods and allow for the normal wait to attendance. Keep rescheduled appointments pending until their outcome is known. Look for operational explanations such as unclear deposits, slow replies or appointment times that do not match the advertised offer.

Use broad administrative cancellation themes in management reporting, not detailed clinical reasons. A higher cancellation count may reflect capacity or communication rather than search quality. Record what changed before deciding that a keyword, agency or landing page caused the difference.

Your monthly cost-review checklist

Bring marketing and reception together around one small report. Review the denominator and the operational context before the headline cost. Make one documented change where possible, and give its enquiry group enough time to mature before judging it.

• Use stable definitions for raw, unique, relevant, booked and attended enquiries.
• Remove tests and duplicates consistently, and show pending outcomes separately.
• Include comparable advertising, agency and handling costs within the stated boundary.
• Review search intent, negative-keyword scope and the actual landing-page promise.
• Check response times, capacity, cancellations and rescheduling before blaming lead quality.
• Verify policy eligibility and tracking consent without exporting patient information.
• Report attribution limits and avoid treating small differences as proven improvements.

FAQ: Should we stop reporting cost per lead?

No. Keep it as an early diagnostic measure with a precise definition. Place it beside administrative relevance, handling effort and mature attendance outcomes. It becomes misleading when it is presented as the clinic's final acquisition cost.

FAQ: Is a higher cost per enquiry always acceptable?

No. It needs to be justified by comparable downstream outcomes and costs, not a vague claim about quality. Consider sample size, time to attendance and capacity. An expensive campaign can still be expensive after every adjustment.

FAQ: Can we upload attendance records to improve bidding?

Do not upload patient or clinical appointment records as a routine marketing tactic. Keep this analysis internal. Any proposed measurement integration needs specialist privacy, security and platform-policy review; a technically available upload option is not permission to use sensitive data.

FAQ: What is the first practical change to make?

Choose one campaign and reconcile its recent enquiries with reception's administrative outcomes in an approved system. Identify the largest preventable mismatch, then address the ad promise, destination or response process. Better definitions often reveal the next useful change before a higher budget does.