A Product Page Trust Bar Cut Revenue per Visitor 10%: Buy Box Case Study
[ SUMMARIZE WITH AI ]
[ FREE CRO TEARDOWN ]
Find the 3 biggest revenue leaks on your store.
Every day a conversion leak goes unfixed, you're paying for traffic that doesn't buy. Get a 5-minute Loom through your PDP, cart, and checkout, with mockups of the fixes. No pitch.
Get My TeardownA product page trust bar lost us 10.2% revenue per visitor. Three short claims under the add to cart button, on the product pages of a premium handbag brand. We ended the test after 19 days and shipped nothing.
Trust bars are one of the most common "safe" changes on Shopify. Icons, three promises, done. This one was well designed, well placed, and still cost money. The version that added more buy box changes on top of the same trust bar came in nearly flat.
The lesson: more info in the buy box is not free. Every line you add near the button is a line the shopper reads right before paying, so each one has to earn its spot.
What We Added to the Buy Box
The test ran on every handbag product page, desktop and mobile. Control + 2 variations, split evenly in Intelligems.
| Version | What changed |
|---|---|
| Control | The existing product page |
| Variation 1 | A trust bar with 3 icons and claims, right under the buttons: where the bags are designed, 14-day returns and exchanges, and a long repair promise |
| Variation 2 | The same trust bar, plus a numeric rating score under the title, a larger price (16px to 20px), and the color swatches swapped for small product image tiles |
The idea was simple. A premium bag is a big purchase. If shoppers see the rating, the price and the colors more clearly, and get reassured on returns and quality at the button, fewer of them should hesitate.
The Scoreboard After 19 Days
Neither variation beat control on revenue per visitor.
| Version | Revenue per visitor vs control | Orders |
|---|---|---|
| Control | baseline | 547 |
| Variation 1 (trust bar only) | -10.2% | 516 |
| Variation 2 (trust bar + buy box changes) | -3.4% | 533 |
Variation 1 was a clear loser. Variation 2 was close to flat. Roughly 500 orders per group is under our 1,000-order target for a confident call, so we read this as a strong signal, not a final law. But a 10% drop with fewer orders in the variation is not a result you keep running and hope for.
The Trust Bar Is the Common Thread
When two variations share one element and both lose, start with that element.
- The trust bar is the only thing both variations have in common
- The variation with only the trust bar lost the most
- Adding the rating score, bigger price and image tiles won most of it back: variation 2 ended about 7.7% ahead of variation 1 on revenue per visitor
So the buy box changes look like they helped, and the trust bar looks like it hurt. We cannot fully separate the two from this test alone, because variation 2 never ran without the bar. That is exactly what the retest is for.
Halfway Through, It Looked Closer
At 12 days, with about 360 to 370 orders per group, the picture was softer:
- Variation 2: -2.1% revenue per visitor, 43% chance to beat control
- Variation 1: -6.3%, 26% chance to beat control
- Average order value within 1% of control for both
- Variation 2 led on desktop and trailed on mobile
The ranking never changed from day 12 to day 19. And mid-test, order value barely moved while order counts did. So the gap looks like it came from how many people bought, not how much they spent. A trust bar under the button touches the decision to buy, not the size of the basket.
Why a Weak Promise Can Hurt More Than No Promise
Our read is that the claims were not the right ones for this purchase. One of them in particular: a 14-day return window.
On a cheap product, "14-day returns" sounds fine. On a premium handbag, it is a hard number placed at the exact moment someone is deciding whether to spend a lot of money. Instead of removing a doubt, it can create one: "only two weeks?"
That is the trap with product page trust bars. Every claim gets read as a promise, and the weakest one sets the tone. A bar with two strong claims and one limiting claim can do worse than no bar at all.
To be clear, this is our hypothesis, not something the test proved line by line. The test tells us the bar as a whole hurt. It does not tell us which of the three claims did the damage. If you want the broader data on badges, see our breakdown on whether trust badges increase conversions.
The Retest: Buy Box Changes, No Trust Bar
We are keeping the part that looked useful and dropping the part that looked harmful:
- Rerun the variation 2 buy box changes on their own: rating score, larger price, image color tiles, and no trust bar
- Make sure it wins on mobile, where variation 2 still trailed mid-test
- If a trust bar comes back, test it with only the strongest claims, and leave the return window out
If you sell high-ticket products, test each claim in your trust bar before you roll it out sitewide. "It's just reassurance" is not a reason to skip the test. For more on what usually works around the buy box, see our Shopify product page optimization guide.
Want tests like this run on your store, losers included? See how our Shopify CRO agency works.
Trust Bar Test: Reader Questions
Do trust bars on the product page increase conversions?
Not always. In this test a 3-claim trust bar under the add to cart button lowered revenue per visitor by 10.2% over 19 days. Whether it helps depends on how strong each claim is for that product and price.
Why would a trust bar lower revenue?
Every claim reads as a promise right before the shopper pays. A limiting claim, like a 14-day return window on a premium handbag, can add doubt instead of removing it. Mid-test, order value stayed within 1% of control, so the loss looks like it came from fewer people buying.
Did the rating score, bigger price and image tiles help?
They looked positive. The variation with those changes plus the same trust bar ended at -3.4%, about 7.7% ahead of the trust bar alone. We are retesting them without the trust bar to confirm.
How many orders did the test have?
547 orders for control, 516 for the trust bar version and 533 for the full buy box version, over 19 days. That is under our 1,000-order target per group, so we treat it as a strong signal, not a final verdict.
What should go in a product page trust bar instead?
Only claims that are clearly strong for your price point, like a long repair or warranty promise. Leave out anything that sounds like a limit. Then test the bar against no bar before you keep it.