Expert authority badge: +17% in week 1, flat (+1.66%) at week 4
Test run byRubenTailored Commerce
Commentary byLiftArchiveOur analysis, not theirsAdding an expert authority badge, similar to a FrontrowMD doctor endorsement, ended flat: conversion was up only 1.66% after four weeks and revenue per visitor did not change. Ruben ran this test. After one week it had looked like a big winner at +17%, which is why the test is worth studying.
- Week 1 conversion rate: +17.35% (2.72% to 3.19%), with a 99.6% probability of beating control. This looked like a clear winner.
- Week 1 revenue per visitor: +9.18% ($4.04 to $4.41), with a range of -8% to +29%.
- Week 4 conversion rate: +1.66% (2.79% to 2.84%), range -4% to +8%, with only a 71.4% probability of beating control. Ruben reads this as noise.
- Week 4 revenue per visitor: flagged as no change ($4.15 vs $4.13). The early revenue gain disappeared.
The change
The variant adds an expert endorsement badge; exact placement was not shown.
Simplified mockup drawn by LiftArchive from the shared screenshots.
Control
The product page as it was, without an authority badge.
Variant
The same page with an authority badge showing an expert or doctor endorsement. The test was named 'Symbols in image', which suggests the badge sat on the product imagery; the design itself was not shown.
Why it likely came out this way
Our reading of the result, not the practitioner's claim.
- early results often swing widely because a few days of traffic don't cover normal weekly patterns or a full mix of shoppers.
- an authority badge may have had a novelty effect that faded, or the first week's visitors may simply have been unusual.
- on this page, shoppers may already have had enough trust signals, so one more badge added little once the data settled.
Caveats: Sample size was not shared and the badge design was not shown. The main lesson is about test length: a 99.6% result after one week reversed to flat after four.
Who ran it
Ruben (@Rubendgrf) 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 add a doctor or expert endorsement badge to the product page, then more visitors will buy because outside authority makes the product's claims more believable.
- Primary metric
- Revenue per visitor
- Guardrails
- Conversion rate
- Average order value
- Cost of the endorsement program
- Who should test it
- Health, wellness and skincare stores where claims need backing and a real expert endorsement exists. Run it for at least two full weeks before calling it, and expect little effect if the page already has strong reviews and proof.
Questions
Do doctor endorsement badges increase conversions?
Not in this test. After four weeks, conversion was up only 1.66% with a 71.4% probability of beating control, and revenue per visitor was unchanged.
How long should an A/B test run?
At least two full weeks to cover weekly swings. Here a test showing +17% at 99.6% probability after one week ended flat after four.
Why did the early result disappear?
Early data is noisy and can reflect an unusual mix of visitors or a novelty effect. Ruben let it keep running and avoided investing in endorsements on a false signal.
Similar tests
@oliverkenyon@oliverkenyon@oliverkenyon@oliverkenyon
@DaveDiederen@DaveDiederen