Which ads produced treatments people paid for?

Aesthetics is one of the few local categories where paid social genuinely outperforms search for lead volume, which makes the reporting particularly misleading. Meta will show you hundreds of cheap inquiries; it cannot show you which of them bought a package.

Ranking by paid treatment value separates the channel that fills a consultation diary from the channel that fills a bank account. They are frequently not the same, and the gap is usually largest for the highest-value treatments.

Elective aesthetic treatment is discretionary, often discussed with a partner, and frequently booked around an event months away. A consultation in March that becomes a package in July is entirely normal and entirely invisible to click attribution.

Matching on the client rather than the session credits the July purchase to the March ad. It also lets you analyze a season that has already finished, which is how you find out whether last summer's campaign actually paid.

Prepaid packages are how most aesthetic clinics smooth their year, and they make the revenue question genuinely ambiguous rather than merely fiddly. A client pays $2,400 in February for six sessions and takes them through to October. There are two defensible dates for that money and two defensible ways to book it, and the clinic's own records probably hold both — the takings on the day the card was charged, and the redemptions spread across the eight months that followed. Neither is wrong, and a report built on one will not agree with a report built on the other.

For attribution the payment date is the better choice, and not for accounting reasons. It sits closest to the decision, which is the thing the advert actually influenced. Booking the revenue on redemption spreads one marketing outcome across eight monthly reports and makes each month's ranking partly an artefact of what clients happened to redeem that month, which nobody chose and no campaign caused. It also delays the signal: a campaign run in February would not be fully assessable until the following winter, by which time it has been renewed or cancelled on a guess.

Whichever a clinic picks, the rule is to pick once and write it down. Switching halfway through a year produces a trend line measuring the export rather than the business, which is precisely the failure this exercise exists to remove. Clinics wanting both views usually export twice into separate imports and compare them, which takes an afternoon and occasionally reveals something worth knowing: a channel that looks strong on takings and unremarkable on delivered treatment is selling packages to people who do not come back to use them, and that is a retention problem wearing a marketing result.

The monthly membership changed this category more than any device did. A client on a recurring plan at $99 or $199 is worth several multiples of a walk-in appointment over a year, attends on a schedule rather than on a whim, and buys retail while they are already in the building. An advert producing members and an advert producing single treatments look almost identical on cost per lead, sit next to each other in the same account, and are not remotely the same business. One of them is buying a subscriber and the other is buying an afternoon.

Judged over thirty days, the membership advert loses and loses badly. Its first month's revenue is a single installment against a full treatment fee from the one-off booking, so on any monthly report it appears to be the weaker buy. Judged over a year, with every installment tracing back to the client who signed up, the ranking inverts and usually by a wide margin. This is the most common reversal in aesthetic reporting and it is structurally invisible to any tool that closes its books monthly, which is all of the ones inside the advertising platforms.

Running it properly is a matter of date range rather than setup. Export twelve months of takings with the client's phone or email on every row, and each recurring charge joins to the same person and therefore to the same original inquiry. Mapping a column that marks membership rows as Product / service lets the report separate recurring revenue from one-off treatment, which answers the question behind the question: is the membership push recruiting new clients, or is it a discount being handed to regulars who were already coming in every eight weeks and paying full price?

Injectables, devices and retail skincare arrive through the same card terminal and behave nothing alike. Tox is priced by unit, repeats on a twelve-week rhythm and is the habit that keeps somebody attached to a clinic. Device treatments are seasonal, expensive, often bought once and heavily influenced by whatever the clinic has just financed. Skincare is small, high-margin and driven almost entirely by whoever is standing at the desk when a client leaves. One revenue total across all three tells a clinic very little about what its advertising actually bought.

The correction is the money rather than a column. Ranking on collected revenue, with the sale count beside it, already shows whether a channel's total is carried by a handful of large device sales or by many repeat injectable visits, because the average value per sale gives it away. A service-line or treatment column would make the split explicit, but what treatment somebody bought is health information, and it does not belong in a file sent to CloseRev, which signs no business associate agreements. Where there is more than one clinic, a location column mapped as Location is the breakdown to use. Nobody could have seen any of this while everything sat inside one number at the top of a dashboard.

Retail deserves to be in the file even though it feels like it belongs to the front desk rather than to marketing. It is genuine money from a client an advert produced, it is usually the most profitable line in the building, and leaving it out understates exactly the channels that bring people back rather than through once. The match does not read what was sold — only that an amount was paid on a date by somebody the inquiry file also knows — so a product name in the export is ignored rather than treated as an error.