<b>The slow page that only hurt the buyers</b>
A retailer's site-wide page speed looked acceptable on average, so performance stayed low on the roadmap for months while conversion quietly stagnated.
The clue came from joining a load-time metric with the user segment. Fast pages were the blog. The product and cart pages, the ones that actually held money, carried heavy scripts and loaded a full second slower for the exact users trying to buy.
They deferred two third-party tags on checkout and lazy-loaded the review widget, ignoring the already-fast blog.
Largest Contentful Paint on the cart fell from 3.4s to 2.1s, and cart-to-purchase conversion rose from 28% to 34%.
The lesson: site-wide speed averages hide the pages that matter. Measure performance where the revenue is, not where the traffic is.
The Pixel Diary
@thepixeldiaries
<b>The slow page that only hurt the buyers</b>
Этот пост опубликован в Telegram-канале The Pixel Diary. Подписаться можно по ссылке: @thepixeldiaries.