--- title: "How often Google rewrites a meta description: three studies compared" url: https://seomadman.com/reports/meta-description-rewrite-rate-compared section: reports published: 2026-06-19T00:00:00.000Z modified: 2026-06-19T00:00:00.000Z author: Adam Hafez topics: ["Technical SEO"] --- # How often Google rewrites a meta description: three studies compared ## The short answer Two studies measured this six weeks apart in 2020. Portent found Google rewrites meta descriptions 71% of the time on mobile and 68% on desktop, across 30,000 keywords. Ahrefs found 62.78%, across 20,000 keywords and 192,656 pages. Search Engine Journal's "over 70%" headline is coverage of Portent's figure. The gap is method and device split, not a real disagreement. ## Key takeaways - Portent's September 2020 study of 30,000 keywords found Google rewrites meta descriptions 71% of the time on mobile and 68% on desktop, via STAT Search Analytics. - Ahrefs' October 2020 study of 20,000 keywords and 192,656 unique pages found a 62.78% rewrite rate, splitting to 59.65% for fat-head keywords and 65.62% for long-tail keywords. - Search Engine Journal's "over 70% of the time" headline is coverage of Portent's own mobile figure, not a third independent dataset. - Ahrefs found Google displays a page's own meta description in the snippet only 37.22% of the time overall, even before counting partial rewrites. - All three figures come from within six weeks of each other in 2020, so the 62.78% to 71% spread reflects different sample composition and device handling, not a trend over time. Ask how often Google uses a page's own meta description and you can get 71%, 68% or 62.78% rewritten, depending on which September 2020 study you open. Portent and Ahrefs ran independent studies six weeks apart, using different tools and different sample compositions, and neither number is wrong. Search Engine Journal's widely cited "over 70% of the time" headline is coverage of Portent's own mobile figure, not a separate measurement. ## The real numbers, study by study Portent's Evan Hall pulled first-page organic results for 30,000 keywords through STAT Search Analytics, split by mobile and desktop, and wrote a script to compare each rendered SERP snippet against the source page's meta description tag. Any snippet text not contained in the meta description tag counted as a rewrite. The result: Google rewrote the description 71% of the time on mobile and 68% of the time on desktop, meaning a site's own wording survived roughly three times in ten. Portent also found rewrite rates ticked up for positions 4 through 6, which the study speculated was Google trying to boost relevance for results below the top three. Ahrefs' Michal Pecánek ran a separate study a month later, analyzing 20,000 keywords and 192,656 unique ranking pages. Ahrefs found a 62.78% overall rewrite rate, splitting to 59.65% for fat-head (high-volume) keywords and 65.62% for long-tail keywords, meaning the more specific a query, the more likely Google was to write its own snippet. Ahrefs also reported that a page's meta description was shown as the snippet, unmodified, only 37.22% of the time overall, a separate number from the rewrite rate since a "not shown" snippet is a different outcome than a partial rewrite. Search Engine Journal's Matt G. Southern covered Portent's study on publication and titled the piece around the mobile figure, "Google Rewrites Meta Descriptions Over 70% of the Time." It is the source most people quoting "70%" are actually citing, whether they know it traces back to Portent's mobile number or not. ## Why Google rewrites a description Neither study tested Google's reasoning directly; both only measured outcomes. For the reasoning, Google's own Search Central documentation is the primary source. It states Google generates its own snippet "when we think it gives users a more accurate description than would be possible purely from the on-page content," and separately warns site owners that descriptions built from "long strings of keywords... don't give users a clear idea of the page's content, and are less likely to be displayed as a snippet." The same page also flags generic, vague description text as a reason Google is less likely to use what a page provides, and recommends unique, page-specific wording instead. That gives three named factors from Google itself, none of them measured against a rewrite rate by either study here: query relevance (Google picks whichever passage on the page best answers the specific search), keyword stuffing, and generic or duplicated description text across pages. A description that fails any of those tests is a plausible target for rewriting, but no published study ties a specific one of these three factors to a specific percentage point, the way the title tag studies isolated H1 match and character length. ## Why we care **A meta description you wrote survives Google's rewriter a minority of the time**, on both studies here, roughly three times in ten on Portent's numbers and closer to four times in ten on Ahrefs'. That is still a reason to write one carefully: Google's own guidance ties the rewrite decision to genericness and keyword stuffing, both things a site controls, even though no study has isolated those levers with a measured before-and-after the way the title tag research has. **Meta descriptions and title tags are rewritten for different reasons at different rates**, and citing one figure for both elements is a common mistake in SEO decks. This site's [title tag rewrite rate study](/studies/title-tag-rewrite-rate-study) found Google rewrites titles 61.6% to 76% of the time, driven mostly by brand-name stripping and H1 mismatch, mechanics that do not apply to a meta description, which has no equivalent on-page heading to match against. Treat the two studies as answering related but separate questions, not the same question asked twice. ## Frequently asked questions ### Which meta description rewrite rate should I cite, 71% or 62.78%? Name the study. Portent's 71% is a mobile-only figure from 30,000 keywords pulled through STAT Search Analytics in September 2020. Ahrefs' 62.78% is a blended figure from 20,000 keywords and 192,656 pages a month later. Neither number is wrong; "Google rewrites descriptions X% of the time" is not a complete claim without the study attached. ### Is the Search Engine Journal "over 70%" figure a third study? No. It is Search Engine Journal's write-up of Portent's own 71% mobile figure, published two weeks after Portent's post. It confirms and amplifies one dataset rather than adding a second independent measurement. ### Why does Google rewrite a meta description at all? Google's own Search Central documentation says it generates a snippet when it believes that gives a more accurate description than the page's own tag would, and separately warns that keyword-stuffed or generic descriptions are less likely to be used as written. Neither study tested Google's stated reasons directly; both only measured the outcome. ### Does this apply the same way to title tags? No, title tags and meta descriptions are rewritten for different reasons at different rates. See this site's separate coverage of [how often Google rewrites a title tag](/studies/title-tag-rewrite-rate-study), which found a 61.6% to 76% range driven mostly by brand-name removal and H1 mismatch, factors these two meta-description studies do not test for. ## Sources 1. [How Often Google Ignores Our Meta Descriptions](https://portent.com/blog/seo/how-often-google-ignores-our-meta-descriptions.htm) - Portent (primary) 2. [How Often Does Google Rewrite Meta Descriptions? (New Data Study)](https://ahrefs.com/blog/meta-description-study/) - Ahrefs (primary) 3. [Google Rewrites Meta Descriptions Over 70% of the Time](https://www.searchenginejournal.com/google-rewrites-meta-descriptions-over-70-of-the-time/382140/) - Search Engine Journal 4. [Control your snippets](https://developers.google.com/search/docs/appearance/snippet) - Google Search Central