Shopify speculation rules and the Core Web Vitals
Conservative, moderate, eager or immediate: which speculation rules sites use, and whether those sites pass LCP, INP and CLS more often.
93.7% of Shopify sites have no speculation rules. Of the 6.3% that do, 98% use moderate, and 88% of those sites pass all three Core Web Vitals, against 88% of all Shopify sites.
93.7% of Shopify sites have no speculation rules. Of the 6.3% that do, 98% use moderate, and 97% of those sites have a good LCP, against 96% of all Shopify sites.
93.7% of Shopify sites have no speculation rules. Of the 6.3% that do, 98% use moderate, and 95% of those sites have a good INP, against 96% of all Shopify sites.
93.7% of Shopify sites have no speculation rules. Of the 6.3% that do, 98% use moderate, and 95% of those sites have a good CLS, against 94% of all Shopify sites.
Core Web Vitals pass rate
Sites with moderate speculation rules pass all three Core Web Vitals most often: 88%, against 88% of all Shopify sites. The biggest gap is LCP, 97% against 96%.
Sites with moderate speculation rules have a good LCP most often: 97%, against 96% of all Shopify sites.
Sites with moderate speculation rules have a good INP most often: 95%, against 96% of all Shopify sites.
Sites with moderate speculation rules have a good CLS most often: 95%, against 94% of all Shopify sites.
Speculation rules usage
6.3% of Shopify sites use speculation rules, and 98% of those use moderate.
Speculation rules per framework
On jQuery, 94% of sites use none speculation rules, and 83% of jQuery sites pass all three Core Web Vitals, against 88% of all sites. On Vue, 92% of sites use none, and 83% of Vue sites pass all three.
On jQuery, 94% of sites use none speculation rules, and 94% of jQuery sites have a good LCP, against 96% of all sites. On Vue, 92% of sites use none, and 95% of Vue sites have a good LCP.
On jQuery, 94% of sites use none speculation rules, and 95% of jQuery sites have a good INP, against 96% of all sites. On Vue, 92% of sites use none, and 94% of Vue sites have a good INP.
On jQuery, 94% of sites use none speculation rules, and 91% of jQuery sites have a good CLS, against 94% of all sites. On Vue, 92% of sites use none, and 91% of Vue sites have a good CLS.
Vue 933 sitesNo rules 92%83%95%94%91%
jQuery 3,153 sitesNo rules 94%83%94%95%91%Speculation rules or the framework
Does the framework explain the speculation rules difference?
On jQuery, 86% of sites with moderate speculation rules pass all three Core Web Vitals, against 83% of jQuery sites with none. Within jQuery, sites with moderate and none pass all three about equally often. On Vue, 84% of sites with moderate speculation rules pass all three Core Web Vitals, against 83% of Vue sites with none. Within Vue, sites with moderate and none pass all three about equally often.
On jQuery, 96% of sites with moderate speculation rules have a good LCP, against 94% of jQuery sites with none. Within jQuery, sites with moderate and none have a good LCP about equally often. On Vue, 99% of sites with moderate speculation rules have a good LCP, against 95% of Vue sites with none. Within Vue, sites with moderate and none have a good LCP about equally often.
On jQuery, 96% of sites with moderate speculation rules have a good INP, against 95% of jQuery sites with none. Within jQuery, sites with moderate and none have a good INP about equally often. On Vue, 91% of sites with moderate speculation rules have a good INP, against 95% of Vue sites with none. Within Vue, sites with moderate and none have a good INP about equally often.
On jQuery, 92% of sites with moderate speculation rules have a good CLS, against 91% of jQuery sites with none. Within jQuery, sites with moderate and none have a good CLS about equally often. On Vue, 93% of sites with moderate speculation rules have a good CLS, against 90% of Vue sites with none. Within Vue, sites with moderate and none have a good CLS about equally often.
How much speculation rules speed up the next page
Speculation rules only speed up the page a visitor opens next, not the page they land on, and Chrome's field data counts both. A prefetch fetches the next page's HTML early, which takes time off its TTFB and its LCP. A prerender also builds the next page before the click, so it can open with an LCP close to zero and its layout shifts happen out of sight. In Shopify's test, moderate cut the median LCP of the next page by 228 ms on desktop compared with conservative, and by 24 ms on phones.
Understanding speculation rules eagerness
Use eager. Sites with eager speculation rules pass all three Core Web Vitals most often, and the theory fits: the earlier Chrome starts, the more of the next page is ready when the visitor clicks. Almost all of those sites are Wix sites, so part of that lead is the platform. The cost is pages loaded that nobody opens, so check that your server can take the extra requests. Conservative waits until the visitor presses a link. Moderate starts after a 200 ms hover on desktop, and on a phone half a second after scrolling stops near a link. Eager starts after a 10 ms hover on desktop, and on a phone once a link has been on screen for 50 ms. Immediate starts as soon as Chrome reads the rule. On WordPress, core uses conservative, which made no visible difference in this data; the wp_speculation_rules_configuration filter switches it to eager.
By Arjen Karel · 8,246 crawled Shopify sites (8,062 with all devices Chrome field data, October 2026). How we measured.