Case study · CrownGenix
One tired ad became a system that keeps finding winners.
Rx hair-loss telehealth, first 60 days on Meta. What the account looked like before, what swipe put into it, and what came out.
Category
Telehealth · Rx
Channel
Meta
Window
First 60 days
Cycles
5

Champion · Cycle 3
Cost per acquisition, cycle 5 against the month before swipe
Click-through rate, from 1.1% to 2.6% on qualified ads
Net-new concepts tested across five cycles, four personas
01 · The starting point
Four ads. One angle. Forty-one days old.
CrownGenix came to us with a familiar account. A small set of template ads built around one message, running to a broad audience, refreshed when someone remembered to. The ads were not bad. They were simply all the same ad, and Meta had stopped finding anyone new to show them to.
Frequency
4.1 on the core audience, climbing week over week
Creative
4 live ads, 1 angle, last new ad 41 days earlier
CPA trend
+22% across the previous six weeks
Learning
No record of which element had ever moved a number
SHOP NOW
Fatigued · Day 41
SHOP NOW
Fatigued · Day 38
02 · Before and after
See the difference
Same persona, same offer, same placement. The left side is the ad the account was running. The right side is what swipe developed, resulting in a 2.4x ROAS improvement
03 · What the model found
Hooks that won.
Hooks that died.
Every cycle, winners were taken apart piece by piece to learn why they won. Three findings shaped the champion.
Won
Price as a daily number
“$1 a day” beat “$32 a month” on hook rate by 2.2x with the same offer.
The pill in a hand
A real pill between two fingers outperformed every lifestyle photo we ran.
Ingredient stacking
Listing all five actives raised click-through among the Skeptic persona by 61%.
Died
Before and after photos
Highest clicks, lowest conversions. Paused in cycle two.
Doctor authority
A clinician headshot read as a stock photo. Flat across all four personas.
Discount framing
“20% off” pulled bargain hunters with a CPA 3x the account average.
Learned
Busy Dad is the account
One persona drove 58% of conversions. Cycles four and five went deeper there.
Comparison wins the Skeptic
The “4 → 5 ingredients” equation is the only concept that moved him.
Night scrollers
Reels after 9pm had the lowest CPA. Budget followed.
04 · Results
Five cycles in.
Click-through climbed every cycle and cost per acquisition fell with it. The dashed line is the real median across our portfolio, so you can see where this account started and where it sits now.
CPA in cycle 5, from $168
CTR, top quartile for the portfolio
Champions found, each from a different persona
Ads older than 30 days still live
05 · Next cycle
The champion is already being taken apart.
Cycle six tests the $1-a-day framing against three new visuals, retires two of the five live concepts, and opens a fourth persona we have not touched yet. The flywheel does not stop because something won. It presses on it.
Put your winners to work.
Join the waitlist. Tell us about your brand and we’ll reach out when a spot opens.