Why title rewrite rate studies disagree: 33% vs 61% vs 76%

Three credible studies measured how often Google rewrites a page’s title tag: 33.4% in 2021, 61.6% in 2022, 76% in 2025. Read as a timeline, that looks like a steady climb. Read against each study’s actual method, it looks like something else: three different rulers measuring the same tape.

What Ahrefs actually measured#

Ahrefs compared the raw HTML title tag against the SERP-displayed title for 953,276 pages ranking in the top 10, using its own Site Audit and Page Explorer index, and published the result in November 2021. Crucially, Ahrefs did not count every difference as a rewrite. A change limited to punctuation, or a brand name Google appended to the end, still counted as a match, and a match required the original word order to survive. Under that definition, Google rewrote 33.4% of titles and used the original 66.6% of the time.

That 33.4% is less than half the 61.6% figure from Zyppy’s 2022 study of 80,959 title tags, covered separately on this site, and less than half the 76% Search Engine Land reported for Q1 2025. Neither of those later studies applied Ahrefs’ tolerant matching rule; both counted any change to the wording as a rewrite. A title tag that only gained an appended brand name would register as unchanged in Ahrefs’ count and as rewritten in Zyppy’s or McAlpin’s.

The H1 finding holds across definitions#

Where the three studies converge matters more than where they diverge. Ahrefs traced 50.76% of its rewritten titles back to the page’s H1 tag, making the H1 the single largest substitute source Google pulled from, ahead of H2 tags at 2.02% and combined H1-plus-H2 sources at 1.31%. Zyppy’s 2022 study, using a stricter rewrite definition and a different sample three years later, also found title-H1 alignment sharply changes rewrite odds: matching the H1 dropped Zyppy’s measured rewrite rate from 61.6% to 20.6%.

Two studies, two different match definitions, two different years, and the same lever comes out on top in both: what a page’s H1 says has more influence over whether Google keeps the title tag as written than almost anything else measured. That agreement is worth more than either study’s headline percentage, because it survives the very methodology difference that makes the headline percentages hard to compare directly.

Length behaves the same way in both studies#

Ahrefs also isolated length as a factor independent of the H1 question: titles wider than 600 pixels were rewritten 46.12% of the time in its later measurement window, up from 29.45% in an earlier one. Zyppy’s later study, measuring by character count rather than pixel width, found the same shape: titles over 70 characters were rewritten 99.9% of the time, versus 39% to 42% for titles in the 51-to-60-character range. Different units, same direction: longer titles get rewritten more, in both a 2021 pixel-based study and a 2022 character-based one.

Why we care#

A rewrite rate is not a fixed fact about Google; it is a number a specific study produced with a specific definition, sample and time window. Treating 33.4%, 61.6% and 76% as three points on one trend line overstates what any single comparison can show, since this piece cannot separate how much of that spread is a real rise in Google’s rewrite behavior from how much is three different definitions of “rewrite” applied to three different samples. What does carry across every definition tested here is the H1 lever and the length lever: align a title tag with its H1, and keep it inside a normal rendered width, and the odds of a rewrite drop under every method these three studies used to measure it.