Essay · Commercial performance
Your Marketing Budget Has a Hole in It
A marketing report can be accurate and still incomplete. Until a principal can reconcile each enquiry to a documented outcome, nobody can say what the budget bought. Here is a one-month exercise to find out before spending more.

In this essay
The meeting
Imagine the monthly marketing meeting. The deck is polished. Traffic is up. Calls and forms have been attributed to campaigns. Cost per enquiry has improved on last month, and the chart showing it slopes in exactly the direction everyone hoped. Nothing on the page is false. The agency has reported, carefully, what its systems can see.
Then you ask one plain question. What happened to them?
The room answers in fragments. The platform attributes forms and calls to particular campaigns. Reception remembers that several people asked about fees, and that one or two sounded keen. The coordinator knows some conversations progressed, because she booked them. The practice-management system knows which consultations were booked. Nobody can join the records without an afternoon of searching across four systems, and nobody has an afternoon.
The dashboard is green. The diary is not. Between them sits the most expensive word in private practice: unknown.
The report ends too early
A conventional acquisition report can legitimately show a great deal: spend, impressions, clicks, calls, form submissions, attributed conversions and cost per recorded enquiry. Better arrangements connect to a CRM, and a good agency earns its fee producing all of it.
But notice what a platform conversion is. In Google Ads, a conversion is an action the advertiser has chosen to count, under counting settings the advertiser also chooses, and the attribution model decides how the credit for it is shared among the ad interactions that came before. That is a sensible way to steer a campaign. It is not a record of what happened inside your practice.
The owner’s question begins where that record ends. A platform conversion is not necessarily a received enquiry; a form can fail and a call can ring out. A received enquiry is not necessarily a useful conversation. A useful conversation is not necessarily a qualified handoff. A handoff is not necessarily a booked or attended consultation. And a consultation is not necessarily treatment, nor should it be.
None of these stages guarantees the next. The point is not to push people along a pipe; it is to know where each of them stands.
There is a line I use with principals: the enquiries you lose never show up on your agency’s report. It is a provocation about visibility, not an allegation of concealment. Unless the downstream systems are reconciled, neither the practice nor its agency can see what happened after the form was submitted. Both are looking at the same hole from opposite sides.
Three systems, each telling the truth
Across the owner-led practices I have worked with, the evidence usually sits in three layers and in several different systems.
Acquisition evidence shows where attention and enquiries were generated: campaigns, keywords, referrals, organic search and social channels.
Communication evidence shows whether each enquiry was received, answered, understood and handed over: the call log, the inbox, the direct messages and the coordinator’s notes.
Practice evidence shows whether an appropriate next step was offered, and whether it was booked, attended, deferred, declined or found unsuitable: the diary and the practice-management system.
Each layer can be accurate on its own terms and still hide the outcome, because no single layer holds the whole journey. Without a defensible connection between them, cost per enquiry can improve while cost per completed consultation deteriorates, or the reverse, and the owner has no way of telling which.
Lost and unknown are not the same thing. A missing booking record is not proof that a patient was lost. It is an unresolved record. She may have booked by phone, chosen to wait or gone elsewhere. Commercial discipline begins by refusing to turn missing evidence into either success or failure. Unknown is not a polite word for lost. It is an instruction to go and find out, starting with the practice’s own records, team notes and systems. A reporting gap is never, on its own, a reason to contact a patient; any follow-up must be lawful, appropriate and consistent with her preferences.
The comforting answer is “we need more leads”
When the consulting diary feels light, the reflex is familiar. Turn the campaigns up. Sometimes that is exactly right.
But the practice cannot know until it separates four problems that look identical from the principal’s chair.
- Insufficient demand. Not enough of the right people are finding the practice.
- Poor-fit demand. People are enquiring, but about the wrong treatments, at the wrong price point or from too far away.
- Failure to receive or progress suitable enquiries. The right people are asking, and the practice is not answering, understanding or handing them on well.
- Failure to record what happened. The practice may be doing perfectly well and simply cannot prove it.
The first is a marketing problem. The second is a positioning or targeting problem. The third is an operational communication problem. The fourth is an ownership and measurement problem.
Buying more media treats all four as the first. If the real problem is the third, extra spend pours more demand through the same gap. If it is the fourth, extra spend may be buying reassurance the practice could have had for nothing. A bigger budget is aimed at the first problem. It cannot diagnose the other three, and it may amplify them. Nobody can say which problem they have until the records are joined.
The Enquiry Reconciliation
This is not a test of response speed; the previous essay dealt with the first reply. It is a reconciliation of the books.
Start with one distinction. A period view shows what happened during a month: the spend, the enquiries that arrived, the consultations that took place. A cohort view follows the enquiries received in that month until their outcomes mature. The reconciliation is a cohort view.
So take one recent calendar month’s enquiries. Create one record for every enquiry the practice can verify it received, across every channel, and keep its original received date. Give each record an internal reference rather than a name, and keep patient-identifiable information inside the practice’s own systems instead of exporting it into a marketing document. Then reconcile seven things for each record.
- Source. Which campaign, referral, organic source or channel produced it, if known?
- Receipt. Did the practice actually receive it, and through which system?
- Useful conversation. Was the person’s question understood and addressed within approved boundaries?
- Handoff. Did the responsible human receive the relevant context?
- Next step. Was an appropriate action offered and assigned to someone?
- Practice outcome. Booked, attended, deferred, declined, redirected, unsuitable or still open?
- Evidence status. Verified, awaiting practice validation or unknown?
Resist the urge to sort every record into won or lost. A patient redirected to a more appropriate clinician is a good outcome. A patient found unsuitable is a correct one. A patient who decides to wait has made a decision, not a mistake. The reconciliation should treat those outcomes with the same respect the practice would show the person in the room.
Reconcile the cohort once at month end, then update the same records at 30 and 60 days, and at 90 where the practice’s normal consideration cycle runs longer. Elective decisions take the time they take. An enquiry still open when the month closes is open, not lost. The platforms keep their own clock: Google Ads credits a conversion to the date of the original ad interaction if it falls within the conversion window, so a month’s reported figures can move after it ends. It is a separate timetable from the patient’s, with the same lesson.
Only then do cost measures mean anything, and only if one rule holds. The numerator and denominator must describe the same money, the same source and the same group of enquiries.
That rule asks for three choices. First, choose the cost scope and name it: paid-media spend alone, or total acquisition cost with agency fees and other costs included and listed. They are different numbers, and sliding between them flatters whichever suits the argument. Second, choose the attributable cohort: only the enquiries that can be defensibly associated with that spend. Organic and referred enquiries stay in the reconciliation, but not in the denominator of a paid-media measure, and source-unknown enquiries stay out of source-specific costs until evidence exists. Third, hold the cohort constant as you move down the chain.
Then three measures become available for a selected campaign or channel.
- Paid-media cost per verified received enquiry. That campaign’s media spend divided by the verified received enquiries attributed to it.
- Paid-media cost per qualified handoff. The same spend divided by the qualified handoffs from the same cohort.
- Paid-media cost per completed consultation. The same spend divided by the consultations that ultimately took place from that cohort, confirmed from the practice’s own records. This is a matured cohort measure, not an instant monthly figure.
One warning deserves its own line. Do not divide April’s advertising spend by every consultation completed in April. Some of those consultations belong to earlier enquiry cohorts; many April enquiries have not yet had time to mature.
Do not put a revenue figure beside an unbooked enquiry. It is not lost revenue. It is a record whose outcome you now know, or know you do not yet know.
The first month is rarely comfortable. In my experience the unknown column is longer than anyone in the meeting expected. That is not failure. It is the first honest number the practice has had.
What a mature agency relationship looks like
None of this makes the agency the villain. Many agencies can only report what they can see. A good one should welcome better downstream evidence, because it is the only way to tell weak demand from weak handling. Without it, the agency tends to be blamed for both.
The constructive arrangement is a division of labour. The agency improves targeting, creative and acquisition. The practice owns response, handoff and clinical pathways. The principal requires a shared, de-identified reconciliation each month, so that both sides learn from outcomes instead of defending their own half of the data.
The meeting then changes its questions. Not “how many leads did we generate?” but these:
- Which sources produced received enquiries?
- Which produced appropriate handoffs?
- Where did context disappear?
- Which records remain unknown?
- What should change next month, in media, content, response design or team workflow?
That is accountability without blame, and a view of the whole commercial journey.
Where the levers meet
In The Exit Discipline, this is several levers working at once. Reaching is the creation and capture of appropriate demand. Converting is the disciplined movement from enquiry to an appropriate next step. Margin is protected when the practice knows whether additional spend is needed before it approves it. And the Scoreboard turns those movements into evidence a principal, or one day a buyer, can actually read. A reconciled month is a small scoreboard.
CareConcierge sits in the middle layer, between acquisition and the practice team. It does not create demand, run advertising or guarantee bookings. Across the channels a practice configures, it provides AI-assisted patient communication: answering from practice-approved content, gathering non-clinical intent, following practice-defined escalation and preparing warm handoffs.
The monthly Conversion Intelligence Report is designed to give the owner a clearer reconciliation layer for the front end of the practice. Where the necessary channels, integrations and practice data are available, it can show progression, stalls, handoffs and evidence that is missing or awaiting validation; bookings, attendance and later outcomes still depend on reliable practice-system data or the practice’s own validation. It does not replace marketing, practice-management or financial records, and it reports unknowns as unknown rather than as estimated success or failure.
Back to the meeting
Return to the room a month later.
The deck is just as polished, and nobody dismisses it. But its numbers now sit as the first ledger rather than the whole story. Beside them is a second sheet: one line per received enquiry in last month’s cohort, with a source, a handoff, an outcome and an evidence status. Some lines say verified. Some say awaiting validation. Fewer say unknown than last month.
The conversation is different. The agency can see which campaigns were associated with appropriate, qualified conversations. The coordinator can see where context went missing. You can see whether the diary is light because demand is thin, or because demand is arriving and not being carried through. You can decide what to spend on the strength of that, rather than on a feeling.
So here is the instruction. Before you add another pound or dollar to the budget, reconcile one month of enquiries. Buy more demand only after you can see what happened to the demand you already created.
If the records show the practice has capacity and is carrying suitable enquiries through, spending more may be exactly the right call. Then it is a decision, not a hope.
Gregory William Gray has spent more than two decades advising owner-led dental, orthodontic and plastic surgery practices in Australia, the United Kingdom and the United States. He is the author of The Exit Discipline and co-founder of CareConcierge Health. More essays and videos are in Inside Private Practice.
Source notes
Google Ads Help: About conversion counting options; About attribution models; About conversion windows. Accessed 27 October 2026. Cited only to explain how Google Ads counts, credits and dates a conversion, not as evidence of campaign effectiveness; other platforms apply their own rules.
Bring a recent month’s enquiry flow, described without patient-identifiable information, to a 30-minute conversation with a founder.