Meridian Store Study
1. What we tested
We tested how easy it is for users to complete a purchase on an e-commerce site. Participants were asked to find a wireless keyboard under $50, apply a discount code (SAVE10), and complete checkout using a saved credit card. Five real users completed the full task in two different versions of the checkout flow — the current design and a new, redesigned version.
2. How it went overall
The task was completed successfully by all participants in the new checkout design, but the current design had serious problems. In the new checkout, users completed the entire task in about 3 minutes on average — all steps went smoothly. In the current checkout, users struggled especially with applying the promo code, where two out of three participants failed or gave up. The new design reduced task time and improved user confidence, with a clear win in both speed and ease.
3. What got in people's way
Hidden promo code field in accordion — a critical flaw
The promo code field was hidden behind an “Order options” accordion, making it invisible to most users. Participants said they didn’t see any place to enter a code, and one directly questioned, “Why would they hide the coupon field inside an accordion? Nobody will find that.” This design flaw affected every session and caused major frustration. It’s not just hard to find — it’s a design mistake that makes users feel misled.
> One participant said: “I don’t see any place for a promo code.”
> Another said: “There’s nothing here that says promo or discount or coupon anywhere.”
This is a critical problem because users don’t know where to look, and the system doesn’t help them find it.
Mismatch between what users thought and what actually happened — a critical failure
Users rated the promo code step as easy (average 4.8 out of 7), but many actually spent a long time trying to find it or gave up. In one session, a participant rated it as 5/7 easy but gave up after 289 seconds — more than four minutes. Another said, “Honestly I would just pay full price at this point. I give up on the code.” This shows a dangerous gap: users don’t realize how hard the step is, and the design fails to show them what’s missing.
This is a critical issue because it means users are misled by the system — they think it’s simple, but it’s not. This leads to frustration, lost sales, and trust issues.
4. Task by task
- Find product: All participants found and added a keyboard under $50 quickly — in about 40 seconds on average. The step went smoothly in every session. Users felt confident and found the product easily.
- Apply promo code: This was the hardest step. In the current design, it took an average of 88 seconds, with one user giving up after 289 seconds. In the new design, it took only 55 seconds and was completed by all users. Users said they had to “poke around” to find the field.
- Checkout and pay: All users completed checkout successfully, but one user (S2) took 121 seconds — more than twice the typical time — and was confused by an error message that said “402” with no explanation. This shows poor error handling and user support.
5. What we recommend
1. Replace the hidden promo code field with a visible, prominent input box — move it to a clear location on the cart page, not behind an accordion. This will help users find it instantly and prevent confusion and frustration. The new design already shows this works — users completed it in half the time.
2. Fix the mismatch between user perception and actual difficulty — users think the promo code step is easy, but it’s not. We must redesign the flow so that users can see the field, know where to enter it, and get feedback if they make a mistake. This will build trust and reduce drop-offs.
3. Improve error messages during checkout — when users get an error like “402,” the system must explain what went wrong and what to do next. Clear, helpful messages prevent confusion and help users recover quickly. The new design already scores much higher in usability (93.8 vs 50.0), showing that these changes make a real difference.
Metrics
2 attempts ran past the axis (longest 4m 49s, against an axis ending at 2m 01s), marked at the right-hand edge. The scale follows the bulk of the attempts so the rest stays readable.
Tasks with fewer than 8 timed attempts show every attempt as a dot rather than a box: quartiles drawn from a handful of measurements would suggest more precision than the data carries.
2 attempts ran past the axis (longest 4m 49s, against an axis ending at 2m 01s), marked at the right-hand edge. The scale follows the bulk of the attempts so the rest stays readable.
Designs with fewer than 8 timed attempts show every attempt as a dot rather than a box. Splitting participants between designs halves the attempts behind each one, so this is the usual case in a moderated A/B study — and the point at which a difference between two designs is easiest to over-read.
Highlight clips
Moments the researcher marked. Play inline.