Where they started
The numbers told the story before we opened the site. Around 40,000 sessions a month, 1.1% conversion, and a cart abandonment rate north of 80%.
The store had eleven apps installed. Reviews, chat, three separate upsell tools, loyalty, a currency switcher, and two analytics scripts that both loaded the same tracking library. Combined, they added 3.2 seconds to first contentful paint on a mid-range Android. Two of the upsell apps were configured to fire on the same trigger and occasionally rendered on top of each other.
Checkout was five steps, with shipping cost appearing at step three. Their own analytics showed the largest single drop-off happened right there — people finding out the shipping price after they'd already committed emotionally.
Collection pages were a product grid and one sentence each. A buying guide on a competitor's blog outranked their own "men's jackets" page for that exact term.
What we built
The app audit came first because it was the cheapest large win. We measured each app's load cost individually, then rebuilt reviews, the currency switcher, and one upsell mechanic directly into the theme. Four apps came out. The two duplicate analytics scripts became one properly configured GTM container.
Checkout went from five steps to three, with shipping shown on the cart page before anyone enters an address, and Apple Pay and Google Pay surfaced at the top rather than buried under the card form. We instrumented every step so the funnel is now visible rather than inferred.
Collection pages got actual content: buying guidance above the grid, subcategory links, sizing and materials information, and an FAQ block. We also set faceted-navigation rules so filter combinations stopped generating thousands of near-duplicate URLs.
