TL;DR: Site speed is a conversion lever first and a ranking signal a distant second. Google and Deloitte measured 30 million mobile sessions across 37 brands and found that a 0.1 second improvement lifted retail conversions 8.4% and average order value 9.2%. Core Web Vitals do feed Google's ranking systems, but Google's own documentation and its Search Advocates describe the effect as small. Optimize speed for the checkout math, not the ranking math.
Why does site speed drive conversions?
Because shoppers leave before the page finishes loading, and the drop-off starts in the first three seconds. Historical Google mobile research found that 53% of mobile visitors abandon a page that takes longer than 3 seconds, and separate Google research found that as load time goes from 1 second to 3 seconds, the probability of a bounce rises 32%. Those figures are older, but the direction still holds, and modern Core Web Vitals now put current numbers on it.
The peer-reviewed literature has documented the same pattern for years. Information systems research established that online shoppers hold a maximum tolerable wait before they abandon a page (Ryan & Valverde, 2006), and a peer-reviewed review of e-commerce systems design found that slower page response times are a major driver of page abandonment, which in turn reduces a merchant's revenue (Ehikioya & Guillemot, 2020).
The clearest number comes from Google and Deloitte's "Milliseconds Make Millions" study, which tracked 30 million mobile sessions across 37 brands. A 0.1 second improvement in mobile load time produced an 8.4% lift in retail conversions and a 9.2% lift in average order value. One tenth of a second.
Vodafone ran the cleaner version of this test: a server-side A/B test on pages that looked and functioned identically, differing only in speed. A 31% improvement in Largest Contentful Paint produced 8% more sales. That's the useful kind of evidence, because nothing else changed.
What does slow load time actually cost?
Run your own numbers, because the answer is usually larger than the fix. Take a store doing 5,000 mobile visitors a day at a 2.5% conversion rate and a $60 average order value. That's 125 orders, or $7,500 a day.
If a hypothetical store saw the same 8.4% relative conversion lift Deloitte observed, that 0.1 second would be roughly 10 additional orders a day. About $630 a day, or $19,000 a month, from one tenth of a second, as an illustrative example rather than a guaranteed result. Their data showed average order value moving too, which we're leaving out of the math to keep the estimate conservative.
Plug in your own traffic, conversion rate, and AOV — our conversion rate value calculator does exactly that math in seconds — before you decide speed work isn't a priority. Most founders we talk to have never done this calculation, which is exactly why the work keeps getting deferred.
Does site speed affect Google rankings?
Yes, but far less than most agencies imply. Google's page experience documentation, updated December 2025, states that "beyond Core Web Vitals, other page experience aspects don't directly help your website rank higher in search results." Core Web Vitals themselves do feed Google's ranking systems, but Google's John Mueller has said they "are not giant factors in ranking, and I doubt you'd see a big drop just because of that."
Google also removed the Page Experience report from Search Console on November 18, 2024, and INP replaced FID as the responsiveness metric in March 2024. Field data updates on a rolling 28-day window, so any improvement you ship takes weeks to appear in the numbers Google actually uses.
One myth to retire while we're here: bounce rate isn't a ranking signal. Google has said repeatedly that it doesn't use it. Fast pages do keep people on site, and that shows up in revenue, not in a ranking bonus.
So treat speed as a conversion project with a small SEO tailwind attached. The 8.4% is the reason to do it. The ranking effect is a bonus you shouldn't build the business case on.
What slows down a Shopify store?
Four things, and most growing DTC brands have all four. Stores in the $500K to $10M range usually outgrow their theme and accumulate code bloat on the way up.
| Performance issue | The fix |
| Unoptimized visuals | Serve WebP, compress before upload, and lazy-load everything below the fold. |
| App bloat | Audit and uninstall unused apps. Defer non-critical JavaScript like chat widgets until after first paint. |
| Heavy theme code | Move off kitchen-sink themes. A lean custom build ships only the code your store uses. |
| Redirect chains | Clean up broken links and collapse redirect chains to cut server response time. |
Uninstalled apps are the one that surprises people. Removing an app from the admin doesn't always remove the code it injected into your theme, so stores carry dead scripts for years. On our Gymshark rebuild, speed work ran alongside the product page redesign for exactly this reason: a faster page and a better page are the same project.
Which speed metrics actually matter?
Three Core Web Vitals thresholds and one business metric. Prioritize Shopify's real-user Core Web Vitals reporting (LCP, INP, and CLS) over chasing a single synthetic speed score, which compresses too much into one number to act on.
- Largest Contentful Paint (LCP): loading. Target under 2.5 seconds.
- Interaction to Next Paint (INP): responsiveness. Target under 200 milliseconds. This replaced First Input Delay in March 2024, so older checklists still naming FID are out of date.
- Cumulative Layout Shift (CLS): visual stability. Target under 0.1.
- Conversion rate: the one that pays. If the vitals improve and conversion doesn't move, the bottleneck was never speed.
Measure field data, not lab scores. Lab tools run on a simulated device on a simulated connection. Field data is what your actual customers experienced, and it's what Google uses.
How does Liquid Lemon handle speed in a 30-day sprint?
It's built into the build, not sold as a separate engagement.
- Days 1-5: baseline audit. Field vitals, high-weight assets, app script inventory.
- Days 6-15: code architecture. Dead app code removed, scripts minified and deferred.
- Days 16-25: media delivery, responsive image implementation, and loading strategy, since Shopify already provides the CDN and automatic image formatting.
- Days 26-30: testing and deployment, targeting LCP under 2 seconds on mobile.
A rebuilt storefront that's slower than the one it replaced isn't a redesign, it's a downgrade, and it happens more than agencies admit. Performance gets engineered in during the build because retrofitting it afterward costs more and works less well. The same logic applies to the CRO mistakes that quietly cap conversion rates.
Build a faster store in 30 days
Liquid Lemon builds fully custom Shopify storefronts in a 30-day sprint at a fixed $9,500, with performance engineered in rather than bolted on.
Run the calculation above on your own numbers first. If the answer is uncomfortable, email hello@liquidlemon.co.
Written by Andrew Zam, Co-Founder of Liquid Lemon
