The Five Hundred Thousand
Enginuity Rx is piloting online sales of a prescription dry-eye treatment through its telehealth platform. Moonbrush ranked 22.4 million adults by behavioral fit, with no health records anywhere in the model, and keyed the pilot’s creative to six sub-cohorts. Results are pending.
Active 2026 pilot
What has been delivered
- ~21.5M
- qualified adults held in reserve for staged expansion
- 0.80
- confidence gate for the in-market cohortScores run from 0 to 1 and rank match strength directionally; they do not forecast conversion rates
- ±7.5 → under ±2.0
- drift against the brand’s own segmentation, before and after recalibrationRoughly ±7.5 points of distributional delta before recalibration, compared percentile by percentile
- 6
- operating sub-cohorts across two macro-cohorts
- 10
- landing-page hero variants, each keyed to a sub-cohortRouted through medical-legal-regulatory review before serving
Who made the pilot
Results pending
No health records in the model. About 21.5M qualified adults are held in reserve for expansion.
Show the numbers as a table
| Adults ranked by behavioral fit | 22.4M |
|---|---|
| Pilot cohort | 500K |
The challenge
Enginuity Rx runs direct-to-consumer acquisition for branded prescription products. For a pilot selling a prescription dry-eye treatment online, it needed an audience it could defend to brand, legal and medical reviewers: people likely to respond, found without any protected health information, and creative that answered each group’s actual hesitation. It also wanted a record, set before any spend, that would show why the pilot performed if it does.
The insight
People who have already bought dry-eye category products have shown, through their own purchases, what the path to treatment looks like. That pattern can be learned once from a stripped seed and then re-expressed purely as behavior: screen and research intensity, wellness purchasing, life-stage markers. Within the qualified audience the hesitations differ sharply. A research-heavy reader decides on mechanism and evidence, someone deciding for the household needs to trust that it is safe and that buying online is sound, and a phone-native user drops off at the first sign of friction. Each needs a different first three seconds on the page.
What we did
- 01
Seed on real purchase behavior
The model started from people who had bought dry-eye category products. Every element of PHI and PII was stripped from that seed before any modeling began.
- 02
Score fit and responsiveness
The seed’s behavioral signature was re-expressed as behavioral analogues, and each person was scored for resemblance to it and for likelihood to engage.
- 03
Correct to the brand’s segmentation
Scored populations were tested against the brand’s own segments, and about ±7.5 points of drift was recalibrated to under ±2.0. A 0.80 confidence gate then set the 500K pilot cohort.
- 04
Key the page to each sub-cohort
Ten landing-page hero variants were written, each tied to one of six sub-cohorts and served by URL parameter on social and programmatic traffic.
- 05
Set the read before launch
Five falsifiable predictions were registered before spend, bound to a measurement plan keyed to completed purchases, and audiences ship in separately tagged confidence bands.
The sub-cohorts
Each sub-cohort gets its own first three seconds on the page.
-
The Continuous Operator
Long-session screen professionals who treat relief as uptime, with the longest daily screen time in the audience.
-
The Stack Optimizer
Systematic self-improvers who audit ingredients and mechanisms and have already worked through over-the-counter options.
-
The Mobile-First Always-On
Phone-native decision-makers who act in short, decisive windows and abandon a process that adds friction.
-
The Diligent Self-Educator
Research-intensive readers who read to the bottom of the page and convert on evidence and mechanism.
-
The Household Steward
Decision-makers for the household who weigh safety, trust in buying online and cost before they commit.
-
The Pragmatic Wellness Investor
People who pay for what demonstrably works, value clarity over novelty and judge health spending by its return.
How it was measured
The pilot is in flight and no results are reported yet. What counts as working was fixed before launch: five predictions registered against the landing page, and a five-part measurement plan keyed to completed purchases. Conversion for the prioritized cohorts will be compared with all other traffic, against pre-launch baseline metrics, in the client’s own reporting at cohort, creative and ad level. Audiences are delivered as separately tagged confidence bands, so the premium 0.80 band can be read on its own and any expansion band is judged on its own performance.
Status
In flight. The audience, the ten cohort-keyed heroes and the measurement plan have been delivered, and results will be read from completed purchases in the client’s reporting. No outcome figures are published until that read is in.
Results describe a specific engagement and are not a guarantee of future results. Platform-reported figures are labeled as such.