TL;DR
Before buying another channel, reconstruct what happened between discovery and an attended PT evaluation. Use the clinic's own records to separate irrelevant traffic, an unclear payment offer, missed intake, unusable evaluation slots and unmeasured outcomes. Buy the repair or test supported by that evidence. “Not much luck” is a starting question, not a channel diagnosis.
Ask for the evidence the next vendor would need
An owner may be considering social advertising because search spending felt disappointing. The useful question is what a different channel would test. If the website generated appropriate inquiries that the clinic could not schedule, changing the traffic source leaves that problem unresolved. If the practice cannot identify any relevant inquiries, inspect campaign targeting and message before concluding that intake failed.
Bring the campaign change log, approved advertising invoices, website destination list and aggregate operational outcomes for a defined review period. Ask the person responsible for the schedule to document the evaluation capacity that actually existed during that period. A calendar full of follow-ups is not evidence that a new evaluation could have been accommodated.
Keep record-level reconciliation inside the clinic's approved systems. The agency needs the authorized findings relevant to its task, not a patient list, diagnoses or treatment notes. Missing source information should remain unknown. A source label invented after the fact cannot recover a lost attribution trail.
Use the channel-to-evaluation diagnostic
This is an original management worksheet. Complete a separate record for each channel and the specific PT service and payment model it advertised. Use actual evidence beside the conclusion rather than rating the channel by memory.
| Boundary | Evidence to inspect | Question for the clinic | Purchasing implication |
|---|---|---|---|
| Discovery to relevant interest | Search terms where available, destination and service promise | Did the campaign attract interest in care we actually offer in our territory? | Repair targeting or offer before repeating the campaign |
| Interest to inquiry | Public contact path and received inquiry categories | Could someone understand cash versus insurance arrangements and reach us? | Clarify the offer or repair the request route |
| Inquiry to contact | Approved intake status summary and staff coverage | Did a responsible person receive and handle the inquiry? | Resolve delivery or coverage before expanding demand |
| Contact to evaluation offer | Available evaluation appointment types and administrative requirements | Could intake offer a usable next step? | Adjust schedule or administrative explanation |
| Offer to attendance | Scheduling dispositions with unknowns preserved | Was the evaluation scheduled, cancelled, missed or attended? | Investigate the specific recorded gap |
| Attendance to financial evidence | Billing team's approved collection summary | Do we have mature collection evidence for the evaluated cohort? | Delay economic conclusions until the evidence is usable |
Do not put a single “marketing failed” label across these boundaries. If the clinic's evidence stops at contact, the next task is to make the later state observable. That finding does not vindicate the campaign or condemn it.
Challenge what the dashboard calls a conversion
Google Ads permits different conversion-counting choices for an action, including counting every conversion after an interaction or limiting the count after an interaction. Ask which action and counting choice produced the reported total. Google Ads conversion-counting documentation.
That setting does not establish whether the underlying event was a request, an evaluation or a follow-up booking. Request the event name, trigger, reporting scope and duplicate-handling rule. If appointment button clicks were labeled new patients, correct the label and preserve the historical limitation. Do not retrofit an attendance claim onto data that never recorded attendance.
Likewise, Google Analytics defines a session as a period of interaction with a site or app. Session counts are not a count of new PT patients. Google's session definition. Use the broader event dictionary to keep the evidence types separate while this diagnostic locates the business constraint.
Decide what changes before the next invoice
Write the result as a conditional repair brief: observed boundary ___; supporting evidence ___; missing evidence ___; responsible role ___; proposed work ___; acceptance check ___. Avoid a conclusion such as “SEO does not work here” when the actual finding is that the clinic cannot reconcile calls with evaluation records.
If the wrong service or geographic audience dominated inquiries, ask for a revised acquisition test with a narrower message. If cash-payment expectations caused confusion, commission an accurate payment explanation. If suitable inquiries reached intake but no evaluation slots existed, operations should decide what capacity can truthfully be promoted before another campaign begins.
If the public path fails, use the booking failure incident method and coordinate a harmless test with intake. Prevent test submissions from reserving unwanted appointments or triggering real outreach. A repaired form should be checked through operational receipt, not merely a success screen.
A new channel can still be a reasonable purchase. Define the unresolved hypothesis, the service and territory, the usable evaluation capacity, the budget owner and the evidence needed to continue or stop. GetPatients has a commercial interest in healthcare visibility services; apply these diagnostic requirements to its proposal too.
Frequently asked questions
Should we try another channel while investigating search?
Only if the clinic can describe what that test will teach and support the inquiries it invites. Keep its evidence separate from the older campaigns. Otherwise the additional activity can make an already unclear record harder to interpret.
What if the previous agency will not supply useful records?
Document the missing accounts, definitions and reporting periods. Recover the practice-controlled evidence that is available and establish clear definitions for future work. Treat the historical conclusion as limited rather than replacing absent evidence with estimates.
Does a low attended-evaluation count prove the ads were poor?
It identifies an outcome that needs explanation. Check the preceding boundaries, including service fit, intake, payment understanding and usable appointments. The count alone does not identify the responsible cause.