Beauty · 2025
Rebuilding a product page around what buyers actually asked
High traffic, high bounce, and a product page written for the brand rather than the buyer. Research found four recurring hesitations; two quarters of testing answered them.
- Client
- A luxury skincare brand
- Industry
- Beauty
- Services
- Conversion Research, Experimentation
- Year
- 2025
- +14.2%
- product page conversion rate
- +9.1%
- revenue per session
- 6 of 23
- tests shipped, the rest documented
The challenge
The brand had built a beautiful product page and a conversion rate well below its category. Sessions were strong and paid acquisition was efficient, but the page did most of its work on brand storytelling and left the practical questions unanswered. Internally there were four competing theories about why, and the team had been redesigning around whichever one was loudest that quarter.
What we did
- 01
Ran on-site polls and eighteen moderated tests with buyers matching the brand's actual customer profile, rather than a generic panel.
- 02
Coded every hesitation raised. Four themes covered the overwhelming majority: ingredient suitability for skin type, what results to expect and by when, whether it suited a routine they already had, and returns.
- 03
Turned each theme into a sequence of tests rather than a single redesign, so we could see which answer mattered and how much.
- 04
Ran everything server-side to avoid flicker on a page where perceived quality is the product, with sample sizes agreed before launch.
The outcome
Twenty-three tests over two quarters, of which six beat control and were rolled into the theme. The winning changes were all informational rather than aesthetic — the brand kept its art direction and answered the questions alongside it. Just as usefully, the programme retired three of the four internal theories with evidence.
Built with
- Shopify Plus
- Intelligems
- GA4
- Contentsquare
- Maze
More work
- Home & furniture2025
Cutting a five-step checkout down to what people needed
A considered purchase with a long deliberation cycle was being handled like an impulse buy. We rebuilt the cart and checkout around delivery certainty.
- checkout completion
- +11.6%
- revenue per visitor
- +6.4%
- steps to purchase
- 5 → 2
checkout completion
revenue per visitor
steps to purchase
- Fashion2024
A mobile site that stopped losing people before it loaded
Four in five sessions were mobile and the page took almost five seconds to render. The fastest conversion win available was not a test at all.
- mobile LCP, 75th percentile
- 4.8s → 1.9s
- mobile conversion rate
- +21%
- repeat purchase rate
- +13%
mobile LCP, 75th percentile
mobile conversion rate
repeat purchase rate
Start here
Find out what your store is actually losing.
Most engagements start with a conversion audit: three to four weeks of research into where revenue leaks in your funnel, and a ranked roadmap of what to test first. The findings are yours to keep, whether you run them with us or not.
- Reply within one business day
- A strategist, not an SDR
- We'll tell you if you're not ready to test