Testing Review Stars Above vs Below the Product Title

Test stars in three positions if your theme allows: above the title as an eyebrow line, directly below the title, and below the price. Segment results by review volume and rating, because placement mostly amplifies whatever the rating already says. A strong rating gains from prominence; a thin or mediocre one loses from it.

Star ratings are the only test element that argues both ways with a straight face. Move them up and you are leading with social proof. Move them up and you are also leading with the weakest number on the page when the product has eleven reviews and a 3.8. Position does not create trust. It sets the volume on whatever trust is already there.

That is why the naive version of this test disappoints. Run it across the whole catalog and the winners and losers cancel: hero products gain, weak products bleed, and the average reports that placement does nothing. Stratify first, and the test starts telling the truth.

Stratify by count and rating before launch

Group products into bands: under 25 reviews, 25 to 200, and over 200, crossed with rating above or below your catalog median. Placement effects concentrate in the extremes. High count, high rating products can carry an eyebrow position. Low count products anywhere near the title invite the shopper to do arithmetic you do not want performed.

If most of your catalog sits in the thin band, the honest test is not position at all. It is presence: stars versus no stars until a product earns them. Many stores quietly threshold star display at a minimum review count for exactly this reason, and that policy decision outperforms any placement tweak.

What to measure, and what to ignore

Primary metric is add to cart rate from the product page, read within strata. Secondary is review section engagement, because moving stars often just relocates clicks that were going to happen. Ignore catalog wide conversion as the headline; it will hide the effect you ran the test to find.

Watch returns and review sentiment in the weeks after for the winning variant if you can. Prominent stars can raise expectations the product then has to meet. A placement that sells more of a borderline product and returns more of it too is not a win at the margin that matters.

Design details that confound the read

Keep the review count visible in both variants, as a bare star row without the count tests a different question. Keep anchor behavior identical: stars that jump to reviews in one variant and do nothing in the other contaminate the comparison with a navigation change.

Mind the mobile fold. Above title on desktop can push the title and price below the first screen on a phone, which taxes every product to pay for the rating line. If the effect splits by device, ship it split. Themes allow it, and the data rarely argues for uniformity.

Frequently asked questions

Do review stars above the title increase conversion?

For products with many reviews and a strong rating, prominence usually helps. For products with few reviews or a mediocre rating, the same prominence hurts. That is why the test must be read in strata rather than as one catalog average.

Should products with very few reviews show stars at all?

Often not until they pass a threshold you set, such as 10 to 25 reviews. A star rating built on three opinions reads as noise and invites doubt. Hiding stars temporarily is a policy many stores adopt after testing presence separately from placement.

What metric decides a star placement test?

Add to cart rate from the product page, segmented by review count and rating band. Catalog wide conversion hides the effect because placement helps strong products and hurts weak ones at the same time.

Should the review count be shown next to the stars?

Yes, in testing and in production. The count is what makes the rating interpretable, and removing it changes the question the test is asking. Keep it identical across variants.