Cutting checkout abandonment by 34% for a skincare brand

The skincare brand came to us selling well on Instagram. The ads were converting, the audience was there, and most of the traffic — 68% of it — arrived on a phone. None of that was broken.

What was broken sat at the very end. People reached the checkout and left. Two out of every three of them. A five-step flow, a forced account signup before you could pay, and a shipping calculator that stalled for several seconds on mobile data while the customer watched a spinner.

We were told for months that our ads were the problem. They were not. People wanted to buy — they just could not get through the last screen.

Where the drop-off actually happened

We put analytics on each step before touching anything. The fall-off was not spread evenly. Nearly half the abandonment happened on the account-creation step, and another large chunk on the shipping screen where the calculator hung. The product and cart pages were fine. The problem was a narrow, fixable stretch at the end — not the store, and not the traffic.

The client's first request was a full replatform to Shopify. It is the obvious move, and the one most agencies would happily sell. We talked them out of it.

Shopify would have cost roughly four times as much, taken three months, and — this is the part that matters — fixed none of the things that were actually losing sales. The checkout was the problem, not the platform. WooCommerce can serve a fast one-page checkout perfectly well. Rebuilding what they already had was cheaper, faster, and aimed at the real fault.

So we rebuilt the checkout in place:

• Collapsed five steps into a single page, with the fields grouped rather than paginated.
• Added guest checkout — despite the client initially wanting every email address up front. We showed them the math: a captured email is worth nothing if the sale never completes.
• Deferred the shipping calculation. Instead of blocking the page while it called the courier API, we let the customer keep filling in details and resolved shipping in the background, falling back to a flat estimate if it was slow.
• Compressed and lazy-loaded the product imagery that was bloating the page on mobile.

Why guest checkout won the argument

The client was not wrong to want emails — they run a genuinely good post-purchase flow. But you cannot email someone who bailed at step three. Guest checkout, with a single honest "create an account?" tick after payment, recovered more addresses than the forced signup ever had, because far more people were reaching the end.

Over the 90 days after launch, checkout abandonment fell by 34%. Mobile conversion — the number that mattered most, given where the traffic came from — rose 41%. Largest contentful paint on the checkout dropped from 4.2 seconds to 1.1.

We were straight with them about one thing: part of that lift is seasonal. The rebuild shipped just ahead of their strongest sales quarter, so some of the gain would have come anyway. We told them to judge it against the same period last year, not the month before. Even on that stricter comparison the improvement held — just not by quite as wide a margin as the headline figure suggests.

None of it required leaving WooCommerce. They kept their platform, their plugins, and roughly eighteen thousand dollars they were about to spend replatforming.