New landing page design: +137% conversion rate, early data
Test run byJosh Thompson
Commentary byLiftArchiveOur analysis, not theirsA new page design beat the existing page with a 137.26% higher conversion rate and 175.64% more revenue per visitor. Josh Thompson ran this split test in Intelligems. The result was still running when shared, with data from August 30 on.
- Conversion rate rose from 2.89% to 6.86% (+137.26%). The range shown was +17% to +381%, so the lift is wide but sits above zero.
- Revenue per visitor rose from $1.15 to $3.16 (+175.64%). The low end of the range is 0%, so the true lift could be far smaller.
- Profit per visitor rose from $0.90 to $2.54 (+181.51%), estimated at +$41,756 per month.
- Average order value rose from $39.67 to $46.09 (+16.18%), but the range of -42% to +141% means this one is not settled.
- Intelligems estimated about 1,013 extra orders and $51,423 more revenue per month.
The change
Whole-page redesign against the old page; the layout of either version was not shared.
Simplified mockup drawn by LiftArchive from the shared screenshots.
Control
The existing page, labelled DDV8 in the test. Its design was not shown.
Variant
A new page design, labelled Josh's Page. The post does not show what the page looks like or what changed.
Why it likely won
Our reading of the result, not the practitioner's claim.
- the size of the lift suggests the old page had a basic problem, such as weak messaging or a hard-to-find buy button.
- a higher order value alongside the conversion gain hints that the new page also steered people to bigger orders, perhaps through bundles or offers.
- without seeing either page, we cannot say which part of the redesign did the work.
Caveats: The test was still live, with data from August 30 and the post dated September 1, so it may cover only a few days. Visitor counts were not shown and the design changes were not shared.
Who ran it
Josh Thompson (@joshtCRO) shared this test on X. The numbers and screenshots are theirs.
View the original post on X
How to run this test on your store
- Hypothesis
- If we replace our current landing page with a redesigned one, then more visitors will buy because the new page explains the offer and next step more clearly.
- Primary metric
- Revenue per visitor
- Guardrails
- Average order value
- Profit per visitor
- Refund rate
- Who should test it
- Stores sending paid traffic to a page that has not been tested in a while. Run it for at least two full weeks, since early results this large often shrink.
Questions
Can a full page redesign beat the original?
It can. Here the new design lifted conversion rate from 2.89% to 6.86%, although the test was still early when shared.
Should I test a redesign or small changes?
A redesign can find big wins, as this one did, but you will not learn which change caused it. Small tests are slower but tell you what works.
How long should a landing page test run?
Longer than this one had when shared, which looked like a few days. Wide ranges like 0% to +662% on revenue per visitor show the data still needed time.
Similar tests
@oliverkenyon@oliverkenyon@oliverkenyon@oliverkenyon
@DaveDiederen
@kasper_dolk