Sarthak UpretiSenior Product Designer
CASE STUDY · GROUP BAYPORT · 2025 – NOW

A checkout that stopped losing people.

Banner Buzz sells custom banner printing in the US at real e-commerce scale. I rebuilt the cart and checkout flow end to end, and rebuilt the homepage. The principle behind both: a buyer who has already decided should never be given a reason to un-decide.

ORGGroup Bayport
ROLESenior Product Designer
YEARS2025 – now
SCOPECart + checkout · Homepage
01 · THE CONTEXT

Custom products make checkout the hardest screen in the store.

A custom banner isn't an SKU pulled from a shelf. Size, material, finishing and artwork all travel with the order, and every one of them is a place where a buyer can stall, second-guess, or abandon. At scale, small leaks are large numbers.

02 · THE CHECKOUT REBUILD

End to end, not patched.

The cart and checkout flow was rebuilt as one continuous decision path: what you configured stays visible, what it costs stays honest, and the next step is always the obvious one. Confirmation work happens before the flow, not inside it, so momentum is never spent re-verifying what the buyer already settled.

BANNER BUZZ · CHECKOUT FLOW
PL.01BANNER BUZZ · CHECKOUT FLOW
03 · THE HOMEPAGE REBUILD

The front door, pointed at the store.

The homepage was rebuilt around getting a buyer with intent to the right product family fast, instead of narrating the brand to them first.

04 · THE NUMBERS

Being confirmed before they go on the site.

The figures below are provisional while the real ones are pulled from analytics, and this site's build physically fails if provisional figures are still marked when it ships. That is how numbers are treated here.

+18%Add-to-cart, after the checkout rebuild
+11%Click-to-cart, after the homepage rebuild