The Median Shopify Speed Score Is 48. When That Costs You Sales.
The median Shopify speed score is 48, and most stores in that range are not losing sales to speed. Here's when the score actually costs you money.
In July 2026 a Reddit user ran Google PageSpeed against 269 live Shopify and DTC stores. Median mobile score: 48. Median LCP: ten seconds. Half of live Shopify stores score under 50. Gymshark's mobile Performance score is 39. Skims is 42. Allbirds is 45. Kylie Cosmetics is 31 and moves nine figures a year. If you tested your store this morning and got a 47, you are the median. That is what a normal Shopify speed score looks like in 2026, once the apps and pixels and product photos are loaded in. Whether that number is actually costing you sales depends on three specific things, and the score itself is barely one of them.
What a 48 actually looks like
A 48 is what a real Shopify store scores once you have added the apps, tracking scripts, page builders, and product media that make a store worth visiting. PageSpeed Insights runs a lab test on a simulated mid-tier Android device on a throttled 4G connection. That simulation is a reasonable model of someone tapping an Instagram ad on a Pixel 6a in a parking lot. It is a poor model of the person on your homepage right now who came from your email newsletter on their office WiFi.
The gap between the 90+ demo scores on theme marketplaces and the 48 you see in your admin is exactly what you would expect when a stripped-down showroom build (thirteen sample products, no reviews app, no email popup, no chat widget, no analytics) meets a real store. One September 2026 benchmark ran 204 headless Lighthouse audits across 17 themes on identical zero-app dev stores. Expanse hit 100 on mobile with a 1.23-second LCP. Broadcast, Symmetry, Sense, and Craft all cleared 99. One premium theme scored 72 before a single app was installed, because its own JavaScript bundle burned 972ms of Total Blocking Time on cold start. That 72 is the ceiling, not the floor, for a store on that theme.
The number moves as you install apps. Community-published data has a store dropping from around 53 with one or two apps to around 38 with sixteen. That is a roughly linear relationship you can predict from the moment you pick a theme. The tool is doing its job.
The three conditions that turn a 48 into lost revenue
You are losing money to a low Shopify speed score when all three of these conditions hold at once. Two out of three usually will not do it.
The first is that a meaningful share of your revenue comes from cold paid mobile traffic. Meta, TikTok, or Google Performance Max clicks landing on a product or collection page from someone who has never seen your brand before. For that visitor, the throttled-mobile Lighthouse simulation is close to what they actually experience. Returning-customer traffic, direct navigation, and branded search are not exposed the same way. If your revenue comes mostly from email, wholesale, or repeat purchases, a 48 is almost never the binding constraint.
The second is that your LCP on your highest-traffic landing page is above roughly four seconds on mobile. Shopify's own enterprise research (28-day CrUX dataset, Jan-Feb 2026) puts real numbers behind the shape most practitioners already suspect: the curve is non-linear. Going from 6 seconds to 2 seconds can double conversion. Going from 55 to 60 on the score is unlikely to move a dollar. The steep part of the curve sits at long LCP. If you are above 4 seconds on the page that most of your paid traffic lands on, you are on the steep part. If your LCP is 2.8 seconds and your score is 48 because your Performance number is dragged down by unused JavaScript on other routes, you are already past it.
The third is that your actual binding constraint is speed and not something else. This is where I have watched merchants burn the most time. A store I looked at earlier this year had improved mobile LCP from 16.9 seconds to 4.7 seconds after three weeks of theme work. Sales did not move. The merchant assumed the speed work had not been enough. When we opened checkout analytics: thirteen sessions reached checkout that week, zero orders completed, against a 30-35% baseline that would predict four or five. Zero from thirteen at the checkout step is a payment gateway signature. The speed work was real. The bottleneck was somewhere else entirely. If you find yourself in that situation and want a second pair of eyes before you spend another week on the Lighthouse tab, the first fifteen minutes with a help1 expert is free and that is usually plenty to spot which constraint is actually binding.
The five failure patterns I keep finding
The same 269-store dataset got a follow-up analysis in August 2026. Five failure patterns appeared in 91% of stores measured: render-blocking scripts, slow first paint, no asset caching, a slow main image (the LCP element), and unused JavaScript from apps. The median SEO score across the same stores was 92. The stores were pouring energy into headings and metadata and structured data (the things they could see and grade themselves on) while the front-end script payload was left alone.
Two of those five patterns produce the biggest LCP wins with the least risk. Fixing the LCP hero image on your homepage and top product pages: making sure it has explicit width and height, loading it as WebP or AVIF, marking it as fetchpriority="high", and never lazy-loading the above-the-fold image. And deferring or removing the app scripts that block first render. A page builder is worth checking specifically. Controlled testing on a clean Dawn theme in August 2026 found PageFly adding 300-600ms to LCP, GemPages adding 300-550ms, Zipify 250-450ms, LayoutHub 200-400ms. A store at 2.5 seconds LCP without a builder can genuinely be at 3.1 seconds on the same page rebuilt in PageFly. If the page is a paid-traffic landing page, that difference is the entire business case.
The practitioner community has a working threshold that Shopify's docs do not endorse: "under 5 seconds is fine." Shopify says LCP under 2.5 seconds is Good, 2.5 to 4.0 is Needs Improvement, above 4 is Poor. Nobody has published a controlled conversion test on Shopify comparing 2.8 seconds against 3.8 seconds on the same store with the same traffic, so the 2.5-second target is authoritative but unproven in the middle range. I recommended chasing a specific score number for a long stretch of 2023 and 2024 and I stopped in early 2025 once I saw enough cases where the score moved and revenue did not. The pass/fail on Core Web Vitals is the frame that actually maps to sales.
Where the tool you are using stops being the right tool
A single PageSpeed Insights run on your storefront will swing your Shopify speed score by 15-20 points across three consecutive tests with no changes made. One merchant documented their Shopify admin score bouncing from 27 to 46 to 32 over six months of no work. That is normal variance for an emulated tool. If you want a reliable lab number, average three runs. If you want to know whether a change you made just now regressed anything, PSI is the only tool that gives a real-time read.
CrUX field data in Google Search Console (Experience > Core Web Vitals) is the number Search actually uses to rank you and the closest thing to what your real customers experience. It is a rolling 28-day window. Shopify's own research found that nearly 80% of Shopify stores pass all Core Web Vitals in CrUX field data, which is the reconciliation with the 48 median. Lab score 48 and CWV-passing are compatible because your real Chrome users skew toward faster devices and WiFi. A store with a 48 lab score and green LCP, INP, and CLS in Search Console has removed all the commercially meaningful speed friction for its actual audience. The gap from 48 to 90 in the lab is not a business problem, as an earlier post on the lab-vs-field gap works through in more detail.
Shopify's Web Performance Dashboard in the admin uses the same CrUX field data with a 28-day lag. It is a smoothing tool, not a monitor. If you break performance today with a bad theme update or a new app install, the dashboard will not flinch for weeks. That gap is real: a 2024 Shopify Community thread documented cases where stores lost sales during high-traffic campaigns while the dashboard showed no problem. Cross-check with PSI immediately after any theme edit or app install. Do not wait for the dashboard to catch up.
And if you are looking at an old Shopify Community thread that says your speed score of 21 is "about average," that thread is from before January 2024 and is talking about a different instrument. The old admin score used a proprietary 10-30 scale. Shopify retired it. The 0-100 PageSpeed number you see today is Google's Lighthouse and reads on a different scale entirely.
What to actually do with a 48
Run PageSpeed Insights once on three URLs: your homepage, your highest-traffic paid landing page, and one bestseller product page. Note the LCP for each. Do the desktop numbers second, if at all. They are usually 25-40 points higher and are not what your paid mobile traffic experiences. Then open Google Search Console under Experience > Core Web Vitals and check the pass/fail status.
If Search Console shows Good on all three metrics and none of your three LCP readings is above 4 seconds, close the tab. Your Shopify speed score is not what is capping your revenue. Go work on the product page copy, the email flow, or the checkout. If Search Console shows Needs Improvement or Poor on LCP or INP, and cold paid mobile is a meaningful share of your traffic, the case for speed work is concrete. Start with the hero image on your top landing page and the app scripts loading in your theme.liquid head. If your zero-app baseline score (duplicate the theme, disable all app embeds, preview URL) is under 60, the theme itself is the ceiling and app cleanup alone will not close the gap. That is when a theme switch enters the conversation.
Skip the specific score number as an outcome measure. Target a CWV pass on your top three URLs. The 269-store dataset had only 9% of stores clearing 90 on mobile. Ninety sits well above the median and the effort to reach it almost never pays for itself.
One last thing
Every merchant I have shown the 48-median figure to has the same first reaction. Relief. Then a follow-up question about what number they should actually be chasing. The follow-up question is the trap. Google is already writing down whether your real customers are having a fast-enough experience, on your Search Console dashboard, in the language it uses to rank you. Your Shopify speed score is a proxy for that answer on a bad day and a distraction from it on a good one.
Still stuck? Talk to an expert.
Our vetted Shopify experts can fix this issue for you in a live session. $39 per session. Your first 15 minutes are free.