0.1 Seconds Moved Conversion 8%. The Speed Evidence Is Better Than You Think

Site speed advice is usually delivered as folklore. The underlying study is better than the folklore. Deloitte Digital, commissioned by Google, analysed four weeks of mobile site data across 37 retail, travel, luxury and lead-generation brands in Europe and the US, comparing behaviour across small differences in load time.

A 0.1 second improvement in mobile load time was associated with an 8.4% increase in retail conversions and a 10.1% increase in travel conversions. Retail consumers also spent around 9% more per order. For lead generation and luxury, where the conversion event is further away, the effect showed up as engagement: page views rose 7% and 8% respectively.

That last point is the one most teams miss. Speed does not only affect the checkout step. It moves every step of the journey — bounce on the landing page, depth of browsing, and completion of the final action. Improving load time is one of the few changes that compounds across the whole funnel rather than shifting a single conversion rate.

Google's own Core Web Vitals thresholds give you the target: Largest Contentful Paint under 2.5 seconds, Interaction to Next Paint under 200 milliseconds, Cumulative Layout Shift under 0.1, measured at the 75th percentile of real users. The 75th percentile matters — lab scores from a fast laptop on office broadband routinely hide the experience of a customer on a phone on a rural connection.

In practice the wins on Irish and UK B2B sites are boringly consistent: uncompressed hero imagery, third-party tag managers loading render-blocking scripts, web fonts without a fallback, and client-side rendering of content that could have been static. Removing one marketing tag has, more than once, delivered a larger improvement than a fortnight of code optimisation.

Treat performance as a commercial metric with an owner and a threshold, not as an engineering nicety. Set the Core Web Vitals targets as a release gate, measure with field data rather than lab scores, and report the number next to conversion rate. The Deloitte data says the two lines move together.