Which channels produced patients who attended?

Mental health services have high no-show and early-dropout rates, and both vary sharply by channel. A campaign producing cheap bookings that do not attend is worse than one producing fewer that do.

Ranking on billed sessions removes the ambiguity, and running the report across a course of care shows which channels produce patients who complete rather than start.

Physician and employer referral produce a large share of patients and are usually treated as unmeasurable, which quietly overstates advertising.

Added to the lead file, with each client tagged by referrer, referral becomes a channel like any other — including which referrers produce clients whose care runs its course.

One client in weekly therapy for eight months is worth more than thirty intake assessments that go nowhere, and the two are indistinguishable on the day the inquiry arrives. Every measurement a practice has access to stops at the intake: the form, the call, sometimes the first booked appointment. The thing that determines whether the practice can pay its clinicians is what happens between session four and session thirty, which nothing in the advertising account will ever observe. The result is a practice buying intakes and hoping for courses, with no way of telling which of its sources supplies the second thing rather than the first.

Sources differ sharply on this and not in the direction people assume. Inquiries from a clinician's recommendation or a professional directory frequently produce longer courses than inquiries from a social advert promising availability this week, and the cost per inquiry usually runs the other way. A practice ranking channels on inquiries received will therefore fund the shorter courses, and will experience the result as a caseload that churns rather than as a marketing error. Owners describe the symptom accurately and diagnose it in the wrong department, usually by asking the clinical team to improve engagement when the problem arrived with the inquiry.

Running the report over six or twelve months rather than one puts the annuity in view. Each session exported as a row against the client's contact detail matches that client and therefore the inquiry that produced them, so the channel ranking becomes revenue by source across the whole course of care rather than clients per dollar. On that comparison a source that looks expensive per new client can turn out to be the cheapest.

Remote sessions changed the geography of this business permanently. A practice licensed across a whole state can see clients three hundred miles away who will never visit the building, while its advertising is still configured around a radius drawn from the parking lot because that is how it was set up. The result is spend concentrated where the office is and demand available where it is not, and no report inside an advertising account can show the difference, because the platform only knows where it delivered an impression.

What it looks like in the data is a source that appears mediocre overall and is in fact excellent in three ZIP codes and poor in the rest. That is actionable and invisible in a blended figure. A location breakdown can show one site carried by marketing and another barely reached by it, and that pattern would not occur to anybody looking at a single national figure. It changes what the next campaign should be told to do, and where.

A location column mapped as Location shows revenue by site, with how much of each traces to a channel. It does not rank channels within each, so the question it answers is whether a site's revenue is being reached by marketing at all, not which channel produces its clients. A delivery-mode or modality column is not one to add: how somebody receives therapy is health information, and CloseRev signs no business associate agreements. Demand from places the practice has not built anything yet shows up too, which is a better footing for a decision about a second site than a hunch about a neighborhood.

Between a full-fee private client, a panel rate negotiated with an insurer, an employer-funded program with its own tariff and a reduced fee agreed on a sliding scale, the same fifty minutes can be worth very different amounts. Practices hold all of these deliberately and for good reasons. What they usually do not hold is any sense of which advertising sources deliver which, and the mix is not even — campaigns that lead on coverage attract people for whom coverage decided the choice.

The practical effect is that two channels with identical cost per inquiry can differ by a third in revenue per session delivered, and the practice experiences this as a vague feeling that one source brings clients who are harder to schedule or slower to pay. Ranking channels on session counts hides it completely. Ranking on collected revenue makes it a number, and once it is a number the decision about where to point the budget stops being a preference. The difference compounds across a course of care, so by session twenty the two channels are not remotely comparable on anything except the headline they were bought on.

Export what was collected rather than what was charged, and nothing describing the fee arrangement: whether somebody pays in full, through an insurer, through an employer program or on a sliding scale is health information, so a fee-type column stays out of the file. A sliding-scale client is not a worse client and the report is not an argument for turning anyone away. Collected revenue per channel already reflects the mix each source delivers, and the assumption of an even spread may turn out to have been wrong for years.