Mobile-First Indexing: What Google Actually Crawls

There’s a version of your website you probably think of as the real one. It’s the one you see when you sit down at your desk, on a wide monitor, with the navigation spread comfortably across the top and every section visible. Then there’s the phone version, which you may think of as the compressed edition, the paperback to the desktop’s hardcover. Mobile-first indexing means Google thinks the opposite. To Google, the phone version is the real one. The desktop version is the variant.

Concretely, this is about which crawler Google sends. Googlebot can visit your site pretending to be a desktop browser or pretending to be a smartphone, and for years now the smartphone crawler is the one whose visit counts. What it sees is what gets indexed, and what gets indexed is what can rank. This transition finished quietly, and most site owners never noticed, because for most sites the two versions carry the same content. But “most” is not “all,” and the exceptions are where the trouble lives.

The failure mode is content that exists on desktop but not on mobile. This happens more innocently than you’d expect. A designer looks at a page on a phone, decides a section feels cramped, and hides it with a CSS rule that applies below a certain screen width. A sidebar full of useful links becomes a hamburger menu, and then someone prunes the menu. A detailed comparison table won’t fit on a small screen, so it’s dropped rather than made scrollable. Each decision is locally reasonable. The sum is a mobile page that’s a subset of the desktop page. Under mobile-first indexing, the missing parts aren’t just missing for phone users. They’re missing, period, as far as search is concerned. You wrote them, you published them, and the index never saw them.

There’s a subtlety worth getting right, because it changes what you should do. Content that’s in the HTML but visually collapsed, say inside an accordion the user taps to expand, is generally fine. Google has said that content hidden this way on mobile is indexed normally, because hiding things behind taps is how small screens work; it’s a legitimate design pattern, not a trick. The dangerous case is content that isn’t in the mobile HTML at all, either because your templates omit it or because it only loads under desktop conditions. The distinction is between folded and absent. Folded is fine. Absent is invisible.

Structured things get lost this way too, not just paragraphs. If your desktop pages have proper titles, meta descriptions, and heading structure, but your mobile template was built later, by different people, from a different base, it may lack them. I’ve seen mobile templates missing the H1 that the desktop version had, or carrying a different, worse title. When my tool GazeSite audits a page, it loads it both ways, as a desktop browser and as a phone, precisely because of this. A finding like a missing H1 or a missing meta description is bad on any page, but on the mobile rendering it’s the finding that actually matters now, because that’s the rendering Google reads.

The deeper mistake underneath all this is treating mobile as an output format rather than as the primary artifact. The order of operations reveals the assumption. If you design the desktop page first and then figure out what survives the shrink to a phone, mobile is where content goes to be cut. If you design the phone page first and then figure out how to use the extra room on a desktop, desktop is where content goes to be elaborated. The second ordering can’t produce the invisible-content bug, because the version Google reads is the version you started from. Mobile-first indexing is Google enforcing, from the outside, a discipline that was already good practice on the inside: most of your visitors are on phones, so the phone experience deserves to be designed, not derived.

How do you check where you stand? The honest way is to look at what the smartphone crawler would see, not what you see. Load your page the way a phone loads it and compare the actual HTML, not the visual impression, against the desktop version. Is everything important present? Are the title, the description, the headings there? Does the main content arrive without requiring interactions a crawler won’t perform? This takes twenty minutes and most people have never done it once, because it never occurs to you that the two versions of your site might disagree. They were supposed to be the same site.

The comforting news is that if your site was built responsively, one HTML document adapting its layout to screen size, you likely have no problem, and that’s exactly why Google favored this transition: responsive sites made it safe. The uncomfortable news is that if you do have this problem, no amount of ordinary SEO effort will show you. You’ll polish pages whose best parts the index has never read. It’s worth an afternoon to be sure. The version of your site you think of as real stopped being the one that counts, and it’s better to learn that from an audit than from a slow, unexplained decline.