NewActiValet is here. Introductory offer: 25% off your first 12 months when you sign up by December 31, 2026.See it →

Google Rewrites Most of Your Titles. Write for That

Jillian Oco, CMO

A peeling shipping label on a crate revealing a different blank label underneath in a dim warehouse.

Google generates the title and snippet it displays from your page's content whenever your tags are weak, outdated, or templated, not from the words in your title tag or meta description. That is why polishing a title tag is not the SEO and AEO fundamental most MSP sites assume. The real fundamental sits one layer down, in heading and body structure.

How often does Google rewrite your title tag?

More often than most marketing teams assume. A 2021 Ahrefs analysis of 953,276 top-ranking pages found Google rewrote the visible search title 33.4 percent of the time, even after counting minor punctuation changes and added brand names as a match rather than a rewrite. A separate, later analysis of 1,469,373 ranking URLs by Portent measured a steeper 63 percent rewrite rate overall, with 67 percent on desktop and 59 percent on mobile.

The gap between those two numbers comes down to method, not disagreement about the underlying behavior. Ahrefs counted small variations as a match, while Portent flagged any difference between the title tag and the displayed title as a rewrite. Either way, a meaningful share of what ranks in the top ten never shows the title the page's owner actually wrote.

Length makes it worse. In the Ahrefs data, titles rendering wider than 600 pixels were rewritten 46.12 percent of the time, up from 29.45 percent for shorter titles. Long titles do not just get truncated. They get replaced.

The same dataset found two other details worth knowing before you audit anything. Nearly 7.4 percent of top-ranking pages had no title tag at all. For pages where the title tag stayed the same across the study period, Google was 33 percent more likely to rewrite the displayed title in September 2021 than in June 2021. Nothing about the page changed in either case. Google's own willingness to override a static tag changed instead, which is the opposite of a stable target to optimize toward.

What happens to your meta description?

Even less of it survives. Google's own documentation says it primarily uses the content on the page to automatically determine the appropriate snippet, and only falls back to the meta description when that element gives a more accurate summary than the page text does. In practice, that discretion gets used constantly.

An Ahrefs review of 192,656 pages across 20,000 keywords found Google rewrote hand-written meta descriptions 62.78 percent of the time, rising to 65.62 percent for long-tail search terms. A separate Portent study of roughly 30,000 client keywords found an even higher first-page rewrite rate, 71 percent on mobile and 68 percent on desktop. The same Ahrefs sample also found that 25.02 percent of top-ranking pages had no meta description to begin with, so a quarter of the field never had a tag for Google to honor or ignore in the first place.

Here is what that looks like side by side.

StudySampleElement checkedRewrite rate
Ahrefs, 2021953,276 pagesTitle tag33.4%
Portent, 20221,469,373 URLsTitle tag63% (67% desktop, 59% mobile)
Ahrefs, 2020192,656 pagesMeta description62.78%
Portent, 2020~30,000 keywordsMeta description71% mobile, 68% desktop

No single number here is the one to memorize. The pattern across four independent studies and two different elements is the fundamental: Google treats your title tag and meta description as a draft it is free to override, not a contract it has to honor.

Why does this matter for AEO as much as SEO?

Because the same substitution habit is what any system has to do when it turns a page into a short answer instead of a ranked blue link. Google's title-link documentation lists the alternate sources it pulls from when a title tag is missing, outdated, inaccurate, or shared across many pages: heading elements, other large or prominent text on the page, og:title tags, anchor text pointing to the page, and WebSite structured data. Those are structural signals, not metadata fields.

Ahrefs' same title tag study checked what filled in when the title tag got ignored. The H1 heading supplied the replacement title in just over half of those cases, with H2 text and combined heading elements accounting for a smaller additional share. A page that buries its real topic below a vague H1, or splits it across several headings that look equally prominent, hands Google nothing clean to extract, whether the output is a search snippet or a generated answer.

This is also where an AEO audit earns its keep. A tool like ActiScore checks whether a page's heading hierarchy, opening paragraph, and title pattern actually agree on one topic, instead of just confirming a title tag exists. That is a different, more useful question than the one most basic SEO checklists ask.

What does Google flag as a low-value title pattern?

A specific one worth knowing by name: micro-boilerplate. Google's documentation describes this as many pages on a site sharing close to the same title, with only a small detail changed, and says it may insert a distinguishing detail itself when that happens.

That pattern is common on MSP sites built from a page template, one page per service, duplicated per city, with the same heading and opening sentence and only the location name swapped. If every variant reads like a placeholder with a city dropped in, Google is explicitly watching for exactly that shape and may decide what the title says on its own, regardless of what the title tag contains.

What does this cost an MSP reselling SEO work?

Time and client trust, mostly. A one-hour title tag pass is an easy line item to sell and an easy one to deliver, which is exactly why it is overused. When a client's rankings do not move after that work, the honest explanation is that you optimized a field Google was always going to override, not that SEO stopped working. That conversation is harder to have after the invoice than before it.

Margin follows the same logic. An audit built around heading structure, topic clarity per page, and template duplication takes longer to deliver once, but it produces findings that still hold up months later, because the underlying page content does not drift the way Google's rewrite behavior does. That is a better foundation for a recurring retainer than a tag checklist a client could run through an AI tool themselves in ten minutes.

How should MSP marketing teams fix this?

Stop treating title tags and meta descriptions as the finish line and start treating headings and opening paragraphs as the real fundamental. A short, practical sequence:

  • Open your highest-traffic pages and check whether the H1 states the actual topic in plain language, not a slogan, before checking the title tag at all.
  • Compare the live search snippet for each page against its meta description. If they rarely match, the fix is page content, not a longer tag.
  • Read your service and location page templates back to back. If two pages differ by nothing but a city name, rewrite the body copy before touching either title.
  • Keep titles unique per page and short, since Ahrefs' rule of thumb from the same study is to keep titles within 60 characters, and wider titles were rewritten far more often than shorter ones.

None of this is a one-time cleanup. New service pages, new locations, and template changes reintroduce the same problems months later, which is why this works better as a recurring audit than a launch-day checklist. If you are deciding which of these checks to package as a resellable service rather than build in-house, the stack builder walks through how a white-labeled audit tool slots into an existing MSP service menu.

Title tags and meta descriptions still matter. Write them well. Just do not mistake them for the thing controlling what shows up in search or in an AI-generated answer. That control sits one layer down, in the headings and opening sentences Google and any answer engine actually read when your tags do not hold up. See the rest of the product lineup built around that same layer, from SEO and AEO audits to the training that gets a marketing team using them correctly. See the full stack to find the piece that matches where your clients' sites are weakest today.