Research-led Shopify Plus rebuild: +21% conversions
Test run bySplitBaseCRO agency
Commentary byLiftArchiveOur analysis, not theirsPrimal Pit Paste saw a 21% rise in conversions after moving to a new Shopify Plus site built from customer research.
SplitBase guided the redesign for the natural deodorant brand, now called Pretty Frank.
The result compares the old and new sites before and after launch, not a split test.
- Conversion rate: +21% site-wide after launch.
- The brand's COO said conversion did not dip after the relaunch, which is a common risk with redesigns.
- Because the platform and design changed at the same time, the lift can't be pinned on one change.
Control
The brand's old BigCommerce store, which the company had outgrown.
Variant
A new Shopify Plus site. Customer research decided the product page layout, the content each page needed and the order in which products and elements appeared.
Why it likely won
Our reading of the result, not the practitioner's claim.
- A platform move forces a redesign anyway, and using research meant the new pages were built around real buyer questions and objections.
- Setting page structure and product order from data likely put the best sellers and key facts where shoppers looked first.
- The new platform may also have brought speed or checkout gains, though the source does not break this out.
Caveats: This combines a platform migration and a full redesign, measured before vs after launch rather than in a split test. No sample size, time frame or confidence level was shared.
Who ran it
SplitBase ran this test for Primal Pit Paste (now Pretty Frank) and published the results (SplitBase case study). The numbers are theirs; the summary and commentary on this page are ours.
Read the original case study on SplitBase →
How to run this test on your store
- Hypothesis
- If we base our new site's page structure and content on customer research, then conversion rate will hold or rise after relaunch because the pages answer the questions buyers actually have.
- Primary metric
- Site-wide conversion rate
- Guardrails
- Revenue per visitor
- Average order value
- Organic search traffic
- Who should test it
- Stores that must migrate platforms or rebuild anyway. If you don't need a new platform, test page changes one at a time on your current site instead.
Questions
Do redesigns usually hurt conversion rates?
They often cause a dip because familiar paths change and untested ideas go live together.
Researching first and launching only proven patterns lowers that risk.
How do I measure a platform migration fairly?
Compare the same weeks year over year and segment by traffic source, since a before-and-after view mixes in seasonal changes. Watch organic search closely too.
What research should come before a redesign?
Analytics on top purchase paths, customer surveys about why people buy, and session recordings or user tests to spot friction.
Similar tests
@oliverkenyon@oliverkenyon@oliverkenyon@oliverkenyon@oliverkenyon
@DaveDiederen