CrawlCheck

Findings · 2026-08-14

The meta tag 1.65 million sites carry and the top 10,000 refuse

MobileOptimized is a Microsoft tag for Internet Explorer Mobile. Adoption across the whole web is climbing. In the top million it is falling, and in the top ten thousand it has never appeared at all.

There is a meta tag called MobileOptimized. Microsoft introduced it to control layout width for markup rendered in Internet Explorer Mobile. That browser is gone. The tag is not.

,

Read as a single number, its adoption looks like a mandate: roughly 1.65 million live sites carry it, and the count is rising steeply. Read by population, it says the opposite.

,

The same tag, four populations

,

Segmented by site rank, the lines disagree with each other. Across the entire web the count climbs. Across the top one million it peaked around 2023 near forty thousand sites and has been declining since. Across the top 100,000, the top 10,000 and the top 1,000 it sits flat against zero for the whole twenty-three-year span the record covers.

,

A technology growing across the web while shrinking among the largest sites is not being adopted. It is being propagated. Nobody with a rendering budget is choosing this tag. It arrives inside themes, page builders and starter templates, and the sites with the expertise to notice have been removing it for years.

,

The jump that gives it away

,

The whole-web line adds several hundred thousand sites in a matter of months in early 2026. Genuine adoption of a twenty-year-old tag for a discontinued browser does not behave like that. A curve of that shape is one platform shipping one default to a very large number of sites at once. We are not naming which, because we have not measured it and a guess dressed as a finding is the failure this site exists to describe.

,

Worth saying plainly: nothing here is bad faith on anyone’s part. A template author includes a line that once did something, nobody removes it, and a decade later it is on a million sites. That is how defaults work.

,

Why the raw count is the trap

,

“1.65 million live sites” reads like evidence that you should add the tag. Segmented, it is a decommissioned signal that the ten thousand most visited sites on the internet have unanimously declined for two decades. Same data, opposite instruction.

,

For contrast, the modern responsive signal — the viewport meta tag — sits above 200 million sites and is present across every rank band. That is what real adoption looks like when you segment it. The shape is flat and high everywhere, not steep in one population and absent in the others.

,

What to do about it

,

If your site declares <meta name="viewport" content="width=device-width, initial-scale=1">, you are done. MobileOptimized and HandheldFriendly target browsers that no longer ship. The related rel="alternate" mobile annotation is worse than useless on a responsive site: it points a crawler at a separate mobile URL you do not have.

,

Removing them changes nothing a visitor sees and nothing a crawler reads. That is the point. They were never doing anything.

,

The general rule

,

This is not really about one Microsoft tag. It is about what a technology-detection profile can and cannot tell you. Those tools list what they can detect, which is not the same as what you should have. A count without a population cannot distinguish a standard from a fossil, and a checklist built on raw counts will confidently recommend both.

,

Before adding anything because a profile says a million sites have it, ask which million. If the answer is “all of them except the ones that know what they are doing”, you have your answer.

,

Adoption figures read from BuiltWith’s published usage statistics on 14 August 2026. We do not republish their table; the segmented counts above are described in our own words and the chart is theirs. BuiltWith’s trend pages are the source for every population figure quoted here, read the same day.

Every figure above came out of this scanner.

Point it at your own domain and see the same measurements, free.

Scan a domain — free

All findings · The dataset · How the dataset works