First-order ROAS mis-prices a specialty clinic's ad spend because it closes the books before the patient decides. Specialty consideration runs 30 to 180 days, and a platform reporting on a 7-day window captures roughly 8 to 38 percent of the consults that spend will eventually produce. The reported figure is not faulty arithmetic. It is arithmetic run too early.
ClinicAds has published a fertility and IVF pillar that walks the consideration window stage by stage, and a med spa post that re-cuts return by treatment line. This post answers a narrower question than either: why the number a specialty clinic reads on day 7 is structurally low, how far off it runs by line, and which metric replaces it. The worked examples here are dermatology, vein, vision, and regenerative rather than fertility.
- A specialty clinic's 7-day reported ROAS captures roughly 8 to 38 percent of the booked consults that spend will eventually produce, depending on the line. A vision practice reading 0.6x at day 7 can be running at 4x measured end to end.
- ClinicAds prices specialty ad spend on cost per qualified consult, in a planning range of $200 to $600 at $4,000 to $15,000 per month of managed media, rather than on first-order ROAS. Those are agency averages, not guarantees.
- Specialty results have to be read by cohort maturity rather than by calendar month. A month whose spend is 20 days old is not comparable to a month whose spend is 160 days old, and stacking both in one dashboard row is the most common version of this error.
Why does first-order ROAS mis-price specialty ad spend?
First-order ROAS mis-prices specialty clinic ad spend because the advertising platform closes its attribution window months before the patient books. Specialty consideration runs 30 to 180 days across dermatology, vein, vision, regenerative, and fertility care. A default window of 7 or 30 days therefore reports a partial cohort as though it were a finished one, and the clinic reads a fraction as if it were a total.
Three separate mechanisms push the reported number down, and they compound rather than overlap. ClinicAds finds all three running in the same specialty account more often than any one of them alone.
- Window truncation: the platform stops counting on day 7 or day 30 while the patient is still inside a 30 to 180 day decision
- Revenue timing: a specialty consult converts to treatment revenue weeks after the consult itself, so even a correctly attributed consult carries a zero revenue value on the day it is counted
- Identity decay: cookie and click-identifier lifetimes expire inside the consideration window, so a patient who returns on day 95 arrives labeled organic or direct
How much of a specialty cohort has booked by day 7?
Between 8 and 38 percent, depending on the line. Dermatology moves fastest because a visible skin concern converts close to the search that started it. Vision and regenerative move slower because the procedure is elective and self-funded, and waiting costs the patient nothing. Fertility is the slowest of the five, which is why it needs its own operating manual rather than a shared benchmark.
The curve below is a ClinicAds planning curve expressed as the share of a cohort's eventual booked consults that have landed by each day. It is an agency average, not a guarantee, and a single clinic's curve will move with price point, payer mix, and metro competition. Read the day 7 column as the fraction of reality a platform-default report is showing.
- A true 4.0x end-to-end return reads as 1.5x at day 7 for dermatology and 0.6x for a vision practice
- A vein cohort adds 33 points of its eventual consults between day 30 and day 90, the largest block any line adds after the reporting window has already closed
- No line in the set has finished by day 30, so a 30-day window is not a safe substitute for a 7-day one
| Specialty line | By day 7 | By day 30 | By day 90 | By day 180 |
|---|---|---|---|---|
| Dermatology, medical and cosmetic | 38% | 66% | 88% | 100% |
| Vein and vascular | 22% | 51% | 84% | 100% |
| Vision, LASIK and refractive | 15% | 42% | 79% | 100% |
| Regenerative, orthobiologic and PRP | 12% | 35% | 72% | 100% |
| Fertility and IVF | 8% | 26% | 64% | 100% |
What decisions break when a clinic acts on the day-7 number?
Four decisions break, and each one makes the following month's number worse. A specialty clinic reading a truncated return concludes the account is unprofitable, then takes the actions a genuinely unprofitable account would justify. Those actions remove precisely the spend that was about to mature, which guarantees the next cohort reports even lower than the one that triggered the cut.
- Killing a campaign in week two that had already produced most of a cohort the clinic will now never see book
- Shifting budget toward the fastest-converting line, which concentrates spend on dermatology and starves the higher-value vision and regenerative lines
- Optimizing the platform bid toward form fills, because a form fill is the only event that lands inside the window
- Ending or renegotiating an agency relationship on a number that has not finished computing
What should a specialty clinic price ad spend on instead?
Cost per qualified consult, held against a cohort maturity curve. ClinicAds works to $200 to $600 per qualified consult on $4,000 to $15,000 per month of managed specialty media, with the window measured end to end at 30 to 180 days rather than on a platform default. Those figures are agency averages, not guarantees, and they describe the operating metric rather than the economic one.
That metric only works if the clinic can tell the platform what happened after the click. The consult outcome has to travel back from the practice management system or the CRM, dated to the original click, which is a plumbing project rather than a reporting setting. The six steps below are the order ClinicAds builds it in.
- 1. Define qualified in writing: the consult showed, met the clinical criteria for the procedure, and was not a duplicate record
- 2. Stamp every lead with source, campaign, and capture date inside the CRM rather than only inside the ad platform
- 3. Extend the platform attribution window to its maximum setting and stop reading the default view entirely
- 4. Send consult-held and treatment-booked events back as offline conversions, dated to the original click rather than to the upload
- 5. Report by cohort month, and never compare a 20-day-old cohort against a 160-day-old one in the same table
- 6. Complete claims-substantiation review before launch, since fertility and regenerative claims draw FTC scrutiny and a paused account measures nothing at all
When is first-order ROAS the right number to use?
First-order ROAS is the right number when the purchase and the click sit inside the same session, which in specialty care applies to very little. The narrow exception is a low-price entry product sold online: a skincare line, a paid teledermatology visit, a refundable vision exam deposit. Those convert within days and can be read honestly on a platform default window.
Everything a specialty clinic actually earns from sits outside that exception. ClinicAds treats first-order ROAS as a creative-testing signal rather than a spend decision. It is useful for ranking which ad earned attention this week, and it is not evidence about whether the account is profitable. The profitability question on a specialty account is answerable at 180 days or not at all.
Does extending the attribution window fix the problem on its own?
It fixes part of it. A longer window recovers the consults that book inside the platform's maximum setting, which on most placements stops well short of 180 days. It does nothing about identity decay, and it does nothing about the revenue that lands weeks after the consult. ClinicAds extends the window and sends offline conversions back from the CRM, because either one alone still under-reports a specialty cohort.
How long before a new specialty account can be judged?
A specialty account produces a readable direction at 90 days and a defensible verdict at 180. At 90 days a dermatology cohort is roughly 88 percent complete and a regenerative cohort is roughly 72 percent complete, which is enough to rank campaigns against each other but not enough to price the account. ClinicAds holds the first 120 days of budget flat for that reason.
What if the clinic cannot connect its practice management system?
Reconcile by hand every month. Export the consult log with source, campaign, and original inquiry date, match it against spend by cohort month in a spreadsheet, and compute cost per qualified consult from that. It is slower than an integration and it produces the same decision, which is the point. A clinic with no connection at all is not measuring specialty performance, it is reading a platform's estimate of it.
Does this apply to dermatology as strongly as to the slower lines?
Less strongly, and it still applies. Dermatology is the fastest of the five lines at roughly 38 percent of its eventual consults booked by day 7, which means a day-7 read still misses about three in five. A dermatology clinic can use a 30-day window as a working signal where a regenerative clinic cannot, but neither should price its media on it.