The biggest names are just as slow
Before the aggregates, the part people do not expect. We seeded the sample with recognizable global brands. Not one of them cracked 50 on mobile. These are companies with nine-figure revenue and dedicated engineering teams.
| Store | Perf | LCP |
|---|---|---|
| Death Wish Coffee | 12 | 15.4s |
| Steve Madden | 25 | 23.3s |
| PopSockets | 27 | 16.5s |
| Dr. Squatch | 27 | 8.3s |
| Kylie Cosmetics | 31 | 18.9s |
| Jeffree Star | 32 | 28.7s |
| ColourPop | 34 | 6.4s |
| Gymshark | 39 | 22.2s |
| Fashion Nova | 41 | 12.6s |
| Pura Vida | 41 | 9.3s |
| Skims | 42 | 7.2s |
| Ruggable | 43 | 5.0s |
| Allbirds | 45 | 15.1s |
| Chubbies | 47 | 8.9s |
If your store scores in the 30s or 40s, you are not behind the industry. You are the industry. That is either depressing or liberating depending on the day.
Median scores: stores nail SEO and fail speed
Across all 269 stores, the four Lighthouse categories split sharply. Stores have clearly invested in the SEO basics (titles, meta, structured data) and largely ignored, or lost the fight on, performance.
Median SEO is 92 and median Performance is 48. The gap is the story: getting found is treated as the job, and being fast enough to keep the visitor is treated as optional. On mobile, where most store traffic lives, that is backwards.
Core Web Vitals: it is a loading-speed problem, not a layout problem
Splitting the field data by Google's Good / Needs-improvement / Poor thresholds shows exactly where stores lose. Layout stability (CLS) is mostly fine. Loading speed is where nearly everyone fails.
88.3% of stores have a poor Largest Contentful Paint, with a median of 10.0s against Google's 2.5-second "good" bar. Meanwhile 82.2% of stores have good CLS. The lesson for store owners: stop obsessing over layout shift and start cutting what blocks the main image and the main thread.
The audits nearly every store fails
Ranked by the share of stores failing each audit. Where we have a plain-English fix guide, the audit links to it.
| Lighthouse audit | Stores failing |
|---|---|
| Network dependency tree | 93.7% |
| First Contentful Paint | 91.4% |
| Use efficient cache lifetimes | 91.4% |
| Largest Contentful Paint | 91.1% |
| Render-blocking requests | 90.7% |
| errors-in-console | 90.3% |
| Time to Interactive | 89.6% |
| Improve image delivery | 88.8% |
| Reduce unused JavaScript | 85.5% |
| Speed Index | 84% |
| Reduce unused CSS | 83.3% |
| Avoid enormous network payloads | 80.7% |
The pattern is consistent: render-blocking resources, unoptimized JavaScript and CSS, and slow image delivery. None of these are exotic. They are the default output of a theme plus a dozen marketing apps, and they are all fixable.
Audit your own store
Paste your URL, get your scores plus a CLAUDE.md fix plan a coding agent can act on.
Methodology
We audited 269 stores with the Google PageSpeed Insights API (Lighthouse 13.4.0), mobile strategy, one run per store, in July 2026. Scores are Lighthouse lab scores; the Core Web Vitals buckets use Google's standard Good / Needs-improvement / Poor thresholds applied to each run's metric values.
The store list is sourced from a public web-performance dataset (the TeamDukaan/performance Shopify list), seeded with recognizable global brands for the comparison above. Composition note: the public list skews toward direct-to-consumer brands and includes a large share of India-market stores, so read the aggregates as "a broad cross-section of real Shopify/DTC storefronts," not a ranking of the global top 269. Lighthouse attached a framework "stack pack" to only 30 of the 269 stores, so this cut is directional, not conclusive. Among the few it did tag, WordPress-detected stores had a higher median performance (73, n=10) than the field, likely because they are smaller catalog/content sites rather than heavy headless storefronts.
Third-party blocking time was excluded: the PageSpeed API did not return a populated value for it on this sample. The full ranked-audit dataset is available as CSV. Anyone can reproduce a single row by running any store URL through lighthouse-md.com or PageSpeed Insights directly.