The challenge
Ads brought people in and carts filled up, but more than half of shoppers vanished at checkout. The team had guessed at pricing, guessed at shipping fees, guessed at trust, and shipped three visual refreshes — nothing moved, because nobody had actually watched a user give up.
How we researched it
We ran think-aloud usability tests with twelve users who matched the customer profile, then cross-read the sessions against funnel data. The break points were painfully specific: forced account creation blocked anyone in a hurry; the sizing chart lived on a separate tab, so people buying boots or packs would not commit; and the return policy was buried on a support page — nobody gambles on high-ticket gear with unclear returns.
What we did
Checkout became guest-first, with account creation moved to after the confirmation page. Sizing and spec information was pulled into the product page and the checkout summary. The return policy was rewritten in plain language and placed directly on the checkout page. Returns themselves became self-serve — request, print the label, track progress — with no email to support required.
Results and the verdict
Checkout completion rose sharply and returns-related support tickets dropped. The verdict is clear-cut: guest checkout was the single biggest lever, self-serve returns came second — and the checkout progress animation we shipped in the same release contributed nothing measurable. It looks nice. Do not expect it to make money.
