I ran a mobile PageSpeed test on a B2B trade site recently. The score came back in the high sixties. Not great, not catastrophic — the kind of number you might see, file under "we'll get to that," and move on.

Then I checked the GA4 data for the same site.

Mobile bounce rate was sitting around forty percent. Desktop was closer to twenty-five. Mobile sessions averaged a couple of minutes. Desktop sessions averaged over twenty. Same site. Same content. Same products.

That order-of-magnitude difference isn't a quirk. It's the real cost of a slow-loading mobile site, and it's the kind of damage that the PageSpeed score alone never makes clear.

This post is about why I think more business owners should be running PageSpeed checks, why the score on its own is misleading, and what to actually do about it when the data confirms there's a problem.

What PageSpeed Insights Actually Measures

Google's PageSpeed Insights gives you a score out of 100 across four categories: Performance, Accessibility, Best Practices, and SEO. The Performance score is built from a set of metrics called Core Web Vitals, the three main ones being Largest Contentful Paint (how long it takes for the main content to render), Cumulative Layout Shift (how much the page jumps around as it loads), and Total Blocking Time (how long the browser is unresponsive while processing scripts).

The thresholds are tight. Largest Contentful Paint needs to come in under 2.5 seconds to score "good." Anything between 2.5 and 4.0 seconds is "needs improvement." Above 4.0 seconds is "poor."

The B2B trade site I tested came in at nearly seven seconds.

That sounds bad until you remember the test runs on a simulated slow mobile connection. Real users on 4G or wifi will do better. So what does it actually mean for your business? On its own, not much. The score is a lab measurement — single-test, simulated network, simulated CPU. It tells you something is technically slow, but not whether anyone is actually suffering because of it.

This is the trap I see most often. A business runs PageSpeed, gets a mediocre score, decides it's not urgent, and moves on. Six months later they're still wondering why mobile traffic isn't converting.

What GA4 Tells You That PageSpeed Cannot

The lab number is theoretical. The GA4 number is real. The pattern on the B2B site I was looking at, segmented by device over the last month, looked roughly like this:

  • Mobile bounce rate sitting in the low forties; desktop around the mid-twenties.
  • Mobile sessions averaging a couple of minutes; desktop sessions averaging more than twenty.
  • Mobile users viewing about two pages per visit; desktop users viewing five or six.
  • Homepage bounce rate on mobile was almost five times what it was on desktop.

That gap doesn't exist because mobile buyers are inherently less interested. It's there because the mobile experience is unusable. They land, they wait, they leave. Desktop users, on a faster connection with more patience, actually stay and browse. Mobile was around a third of total traffic on this site, so this isn't a marginal segment.

When you put the lab data and the field data together, the story becomes clear. The 6.9-second LCP is not a theoretical issue. It's actively costing this business mobile customers every day.

Why the B2B Context Makes It Worse

I want to dwell on this because it's the part most articles miss.

Conventional wisdom says B2B buyers do their research on desktop, so mobile speed doesn't matter as much. The data on that site says the opposite.

Trade buyers use mobile differently to consumers. They're not browsing for fun in the evening. They're checking suppliers between meetings. Looking up specs on a site visit. Pulling up pricing while their boss asks them a question. Mobile is the first-impression channel — the moment a buyer decides whether you're worth coming back to on desktop later.

If your mobile site is slow, that buyer never makes it to the desktop deep-dive. They've already moved to the next supplier in the search results.

The economics of this matter too. A trade account can be worth several thousand pounds a year. Losing one mobile B2B buyer to a slow page isn't comparable to losing one consumer browsing for a £30 cushion. The cost-per-bounced-visitor is far higher in B2B. If you're a B2B business owner reading this and your mobile bounce rate is dramatically worse than desktop, the financial damage is real, even if it's hidden in your analytics rather than appearing on your P&L.

The Hidden Tax: Quality Score in Google Ads

There's a second cost most people don't connect.

Google Ads Quality Score factors in landing page experience, and page speed is part of that calculation. A slow mobile site doesn't just lose you direct conversions — it inflates what you pay per click for the traffic you do get. Every paid mobile click costs more than it should because Google has decided your landing page experience is poor.

I've seen this play out repeatedly. A client tightens up their landing pages, page speed improves, and three or four weeks later the cost-per-click on their mobile campaigns drops without anyone touching the bidding. The fix paid for itself in ad spend before any conversion improvement landed.

So when I'm advocating for mobile speed work, it's not just about user experience. It's about not paying the page-speed tax in your ad account every month.

What to Actually Do About It

If you've read this far and you're wondering what to do next, here's the order I'd suggest.

First, run PageSpeed Insights on your homepage, your top product or service page, and your contact or registration page. Use the mobile setting. Note the LCP number more than the overall score.

Second, open GA4 and segment your bounce rate, average session duration, and pages-per-session by device category for the last 30 days. If your mobile metrics are dramatically worse than desktop, you have a real problem — regardless of what PageSpeed told you.

Third, if both data sources agree there's an issue, the standard fixes are well understood. Optimise the largest above-the-fold image (resize it, compress it, add a preload tag). Strip out unused JavaScript and CSS that's loading on every page. Make sure your hosting and CDN are configured properly for the markets where your buyers actually are.

None of these are expensive. They need a developer afternoon, not a website rebuild. The trick is recognising when they're worth doing — and that recognition only comes when you stop looking at the PageSpeed score in isolation and start cross-checking it with what real users are actually doing on your site.

The Wider Point

Most marketing data is lab data. PageSpeed is lab data. SEO audit tools mostly produce lab data. Even Google Ads keyword planner gives you simulated estimates. None of it tells you what real customers are doing on your real site.

GA4 is one of the only tools that does, and it's free. Most businesses I work with under-use it dramatically. If you take one thing from this post, let it be that the real story of your customer experience is sitting in GA4 right now, and almost nobody is looking at it through a device-segmentation lens.

The lab number gives you the diagnosis. The field data tells you whether anyone's actually sick.

Go look.