Why the product wasn’t failing.
The message was.
The Role
Owned insights data reporting, Paid & Creative Strategy, PDP rebuild using Replo, and testing frameworks.
The Problem
The Filtered Showerhead wasn’t scaling.
Ads were getting watched, but not acted on.
Spend collapsed into a small number of “safe” creatives.
Testing kept stalling.
On paper, the product was solid.
In reality, it was being sold to the wrong identity.
People actually buying weren’t looking for function.
They were looking for how it felt.
The Insight
I spotted the gap by combining three signals:
CX tickets repeatedly referenced skin, hair, irritation, and routine
My own first-hand experience installing and using the product
Paid ad data showing strong hook rates but weak conversion signals
People weren’t buying hardware.They were buying a beauty and comfort upgrade.
What I did.
01
Rebuilt the PDP from function-led to beauty led
02
Fixed testing structure so Meta could actually explore (CBO testing, CPR off)
03
Designed a persona-led testing system (beauty, sensitive skin, comfort, performance)
Products don’t fail. Positioning does.
Meta responds to identity faster than explanation.
Results
Meta consistently favoured beauty-led and routine-led personas
CTR and engagement improved relative to function-led ads
Spend unlocked into fewer, clearer persona winners instead of clustering randomly
Shower testing delivered 185 results at a average £50 CPA during structured testing
