Apple, Meta and Microsoft Are Among the World's Most Valuable Companies. Their Websites Do Not Pass Core Web Vitals.

Apple, Meta and Microsoft sit among the most valuable companies ever built, worth trillions of dollars between them and staffed by some of the best engineers alive. Their public websites do not pass Core Web Vitals. In Google's own field data, Apple's site is measured at 3.8 seconds to load on mobile, Meta's at 4.4, and Microsoft posts a 1,875-millisecond delay after a tap, the slowest interaction reading in the entire group. Of the world's 50 most valuable public companies, just 18 pass.

This is not a cosmetic finding. Speed shapes whether visitors stay, whether they convert, and how a brand ranks in search. Google's research links slower pages to sharply higher abandonment, retail studies tie fractions of a second to real revenue, and Core Web Vitals has been a Google ranking signal since 2021. When the wealthiest companies in the world deliver a mobile web experience below Google's good thresholds, it says something useful: budget is not what stands between a site and a fast one.

Every reading here comes from the Chrome User Experience Report, Google's public record of what real Chrome users experienced on real devices and networks, and the same family of field signals Google's ranking systems use. Ranked by market value, the picture is not the one you would guess, and the companies that pass are not the ones you would expect.

We compiled the 50 most valuable public companies by market value and queried the Chrome User Experience Report for each company's registrable domain, taking its busiest origin, usually a corporate, investor or retail site rather than a product app, measured on mobile at the 75th percentile. A missing INP counts as failing.

A company passes only when all three metrics land in the good range at the 75th percentile: Largest Contentful Paint under 2.5 seconds for loading, Interaction to Next Paint under 200 milliseconds for responsiveness, and Cumulative Layout Shift under 0.1 for visual stability. Every value below is a real reading in Google's public dataset, and the 18 companies that pass show it is an entirely reachable bar for a flagship corporate site.

There is a comfortable assumption that scale buys speed, that a company worth a trillion dollars with thousands of engineers must run a fast website. Google's field data reads differently. Of the 50 most valuable companies, 18 pass on mobile, so nearly two-thirds of the biggest firms by value currently deliver a mobile web experience that sits below Google's good thresholds. Budget is clearly not the whole story.

The Big Tech results are the most striking part. Apple's site is measured at 3.8 seconds at the 75th percentile, Meta's at 4.4, and Nvidia, Tesla, Adobe and Salesforce each miss at least one threshold. Microsoft posts the slowest single reading in the group, a 1,875-millisecond Interaction to Next Paint against a 200 ms good bar, with anything over 500 ms classed as poor. For the company that builds so much of the world's productivity software, that is a large gap between the site and the thresholds, and it is exactly the kind of gap engineering time can close.

The slowest loaders in the set are not the tech giants at all but telecoms and industrials: Comcast at 5.9 seconds, Qualcomm at 5.7, IBM and JPMorgan Chase at 4.8, with Meta, Tesla and Apple close behind. These are marketing and corporate sites carrying heavy front-ends, not lean product surfaces.

The tech companies that do pass are largely the ones whose business is the web itself: Google is measured at 1.1 seconds, Amazon at 1.6, and Netflix passes comfortably. The split inside tech is really platform-native versus product-and-marketing, and the platform-native giants win it. The rest of the passing group is a surprise only if you expected the winners to be glamorous. They are the old-economy names: ExxonMobil, Merck, Johnson & Johnson, AbbVie, Caterpillar, PepsiCo, Intel, Costco, Home Depot, Nike, Goldman Sachs, McDonald's and Boeing.

What they tend to share is restraint. Corporate, investor and retail sites that behave more like documents, light and quick, without the heavy interactive marketing machinery that weighs on the tech giants' pages. A plainer site turns out to be a performance advantage: the passers carry less, and it shows in every metric.

It would be easy to read these numbers as a knock on the companies, and that is not what the data supports. A homepage for one of these firms is rarely a simple page. It carries rich, interactive, marketing-driven experiences: marketing and analytics stacks, tag managers, consent tools, personalization, A/B testing, and a long list of third-party scripts that each add main-thread work no single team fully controls. Many run at global scale across dozens of markets and languages, on content platforms chosen years ago that are expensive to replace, and the homepage is usually built for brand and campaign flexibility first, which pulls against a tight performance budget.

That makes speed a constant negotiation between many teams, and the field data captures where that negotiation currently sits. Microsoft's tap delay is not a hardware limit. It is main-thread JavaScript, the kind of thing a focused effort can bring down. The full table shows the whole set at once, ranked by market value, with a 36.0% pass rate and the three metrics broken out.

All 50 companies, ranked by market value, mobile, 75th percentile, May 2026 CrUX. Pass rate 36.0% (18 of 50). Good LCP 42.0%, good INP 55.1%, good CLS 80.0%. A missing INP counts as failing.

The 36% pass rate settles the budget question. When ExxonMobil clears the bar at 1.9 seconds and Johnson & Johnson at 1.7 while Apple sits at 3.8 and Meta at 4.4, the difference is not money or engineering talent. It is how much the page insists on doing before a visitor can use it, and the old-economy passers post better field numbers because their corporate and investor pages carry less.

The metric that fails points to the work that fixes it. Microsoft's 1,875-millisecond tap delay against a 200 ms good bar is main-thread JavaScript a focused effort can bring down, not a hardware limit. Comcast at 5.9 seconds and Qualcomm at 5.7 are slow starts and heavy front-ends, a loading problem that lives in what the page pulls in before its hero paints. Reading your own failing metric tells you whether the job is the server or the scripts before you rebuild anything.

The distance between the 18 passers and the 32 that miss is a competitive opening. Google's field data grades a trillion-dollar homepage on the same three thresholds it applies to a local business, so a fast site is a differentiator budget cannot buy back. Ship a lighter page and defer the scripts that keep the main thread busy, and a flagship site can match Google's 1.1 seconds and Amazon's 1.6 on the numbers search actually ranks.