- Abbandono Carrello, Core Web Vitals 2026, ingegneria ecommerce, Protocollo IFG eCommerce, ROI Shopify, Shopify Performance
- Francesco Guiducci
Shopify PageSpeed: Why 99/100 Is Not the Goal
Direct answer: a PageSpeed score of 99 or 100 does not eliminate cart abandonment and does not guarantee conversions. Lighthouse is a lab test. I use it as diagnostic evidence and compare it with real Core Web Vitals and the behavior of the actual template and device.
PageSpeed is not real traffic
The test simulates defined conditions and produces a synthetic score. It can reveal opportunities but cannot replace field data from real users.
TBT and INP are different
Total Blocking Time is a lab metric that highlights main-thread blocking JavaScript. INP is a Core Web Vital measured across interactions. Lower TBT can help, but there is no fixed numerical relationship between TBT, INP and conversion.
The thresholds I use
Google's “good” thresholds are LCP within 2.5 s, INP within 200 ms and CLS within 0.1 at the 75th percentile. I do not invent a private threshold and turn it into a sales claim.
What I check on Shopify
- LCP image and resources;
- theme and app JavaScript;
- CSS, fonts and layout shifts;
- Liquid rendering;
- assets loaded where they are not needed;
- mobile regressions after changes.
Do not break the store for a score
Blindly deferring scripts can break variants, analytics, popups, apps or cart behavior. Every optimization needs a functional regression test, not only a better score screenshot.
Performance and business
A more responsive page removes a real friction, but revenue also depends on product, price, traffic, trust and checkout. I do not assign generic ROI tables to milliseconds.
Updated September 3, 2026.
