What's a good Lighthouse score for Shopify? Real benchmarks

A merchant emails me with a screenshot. “Lighthouse says 42. Is that bad?”

The honest answer: it depends. 42 is below average overall, but for an active Shopify store with 20+ apps and a real product catalog, 42 might be the best you can get without surgery on the theme. For a brand-new store with three products and no apps, 42 is alarming.

This is the post I send when someone asks. Real benchmarks, what’s normal, and what’s actually worth chasing.

Sub-post of the Shopify Speed Optimization Guide.

TL;DR

• Mobile Lighthouse is the only one that matters. Don’t celebrate a 95 desktop score.
• For active Shopify stores: 50–70 mobile is the realistic target. Above 70 takes serious work; above 90 means you’ve cut a lot of apps.
• Lighthouse is a lab score — synthetic, single-device, single-network. Real-user data (CrUX in PageSpeed Insights, or Search Console → Core Web Vitals) is what affects rankings.
• Don’t optimize for the score — optimize for the metric. LCP, INP, CLS in the “Good” range on real users is the actual goal.

What Lighthouse actually measures

Lighthouse runs your page in a simulated mobile browser, on a simulated 4G connection, and measures four things rolled up into the Performance score:

• First Contentful Paint (FCP) — first time anything shows up
• Largest Contentful Paint (LCP) — biggest content element painted
• Total Blocking Time (TBT) — how much the main thread is blocked (proxy for INP)
• Cumulative Layout Shift (CLS) — how much stuff jumps around
• Speed Index — visual progress curve

The Performance score is a weighted average. The biggest weights are LCP (25%) and TBT (30%), which is why those are the two metrics speed work usually targets first.

What “good” looks like for Shopify

Lighthouse’s official thresholds are universal — they don’t know you’re on Shopify.

Score	Lighthouse classification
90–100	Good
50–89	Needs improvement
0–49	Poor

The reality on Shopify, from the stores I’ve audited:

Store profile	Typical mobile Lighthouse
Brand-new, default Dawn theme, 0 apps	75–95
Small store, OS 2.0 theme, 5–10 essential apps	55–75
Active store, themed-up, 15–25 apps, GTM, chat	35–55
Older OS 1.0 theme, heavy customization	20–40

If you’re in the third bucket and scoring 45, you’re roughly average. If you’re in the same bucket scoring 25, something specific is broken — usually one app or one theme section.

Why the score doesn’t matter (much)

The score is a single number summarizing a complex page load. The trap is optimizing for the score instead of the experience.

Two things to know:

1. The score is variable. Run Lighthouse three times on the same page, you’ll get three different scores — sometimes within ±10 points. The simulated network and CPU throttle introduce variance. Don’t celebrate or panic over single-run swings.

2. The score is lab-only. It doesn’t reflect your actual users on their actual phones with their actual networks. The number Google ranks on is CrUX — the Chrome User Experience Report — which aggregates real Chrome users’ visits over the past 28 days.

Always check both:

• Lighthouse = lab, directional, useful for finding issues
• CrUX (in PageSpeed Insights, top of the report) = field, ranking-relevant, reality

What’s realistic to chase

For a stocked Shopify store, here’s what I’d target — in this order:

1. Mobile Lighthouse 50+. Below this, something is genuinely broken; you can usually find it and fix it.
2. CrUX “Good” on LCP (< 2.5s mobile). This is the single biggest ranking signal in CWV. Most stores can hit this.
3. CrUX “Good” on INP (< 200ms mobile). Harder than LCP — driven by JavaScript work, often from apps. Reaching “Good” usually means cutting or deferring scripts.
4. CrUX “Good” on CLS (< 0.1). Easiest of the three — it’s about reserving space for images and ads so the page doesn’t jump.
5. Mobile Lighthouse 70+. Possible but takes work. You’re cutting apps, refactoring sections, maybe touching the theme code.
6. Mobile Lighthouse 90+. Possible but hard. You’ve removed most apps, optimized every image, self-hosted fonts, stripped third-party scripts. For most active stores, this isn’t worth the trade-off.

What’s not worth chasing

• Desktop Lighthouse 100. Easy to hit, doesn’t matter. Google indexes mobile.
• Lighthouse Best Practices / SEO scores. Often mislabeled as “speed”. They’re separate audits and have separate fixes. Don’t conflate them.
• The score itself, if your CrUX is already “Good”. If real users are having a good experience and Search Console says you’re in the green, the lab score is academic.
• Score parity with competitors. They might have fewer apps, less merchandise, simpler homepages. Apples to oranges.

A simple decision rule

When a merchant asks me whether their score is bad, my answer follows this rule:

If CrUX (real users) is "Good" on all 3 CWV → you're fine, ignore Lighthouse
If CrUX is "Needs Improvement" → Lighthouse score is a useful diagnostic
If CrUX is "Poor" → fix it, regardless of Lighthouse
If CrUX has no data (low traffic) → use Lighthouse as the proxy

For a store with low real-user traffic — say, a new launch — you only have lab data to go on. Run Lighthouse three times, average it, and use that as your baseline.

What to actually do with your number

1. Run PageSpeed Insights on three pages: home, a collection, a product. Use mobile.
2. Look at CrUX first (top of the report). That’s reality.
3. Look at Lighthouse second. Note which Core Web Vital is in red.
4. Cross-reference with the diagnostic guide. Most slow stores trace back to one of six causes.
5. Fix the biggest one. Re-measure.

If you’re stuck — score is low, you’ve cut what you can, CrUX is still “Needs Improvement” — that’s where a focused speed optimization engagement usually pays for itself. Most engagements move mobile CWV into “Good” within 2–4 weeks.

---

Want a 5-minute speed read on your store? Book a free intro call. Send your URL beforehand; I’ll run live Lighthouse + CrUX checks and tell you what’s realistic to chase given your stack.