Technical SEO · September 9, 2026 · 7 min read
Google Rewrote Your Title Tag: Why It Happens and How to Get Yours Back
Title tag rewriting is Google's default behaviour rather than an occasional glitch, and roughly a third of results show something other than what the page specified.
By FluxWriter Team
Title tag rewriting is Google's default behaviour rather than an occasional glitch, and roughly a third of results show something other than what the page specified. Most owners discover this while diagnosing a click-through drop and assume something broke. Nothing broke — the system decided something else on your page described it better than your title did. This covers why it triggers, how to check it, and the edits that reliably win the tag back.
Why Rewriting Happens at All
Google generates the title link from the page itself, not from the query someone typed. That single fact explains most of the confusing behaviour around this.
The tag you write is the strongest candidate, not the final answer. The system also considers your H1, anchor text pointing at the page, other prominent on-page headings, and the site name. When your tag looks like a poor description of the page itself, one of those alternatives gets used instead.
Rewrites became far more common after the 2021 title generation change, and the rate has stayed high since. Treat it as normal operating behaviour. The goal is not preventing rewrites — it is making your version the obvious best candidate.
The Four Triggers That Cause Most Rewrites
Rewrites are not random, and the causes cluster tightly. Four patterns account for the large majority.
Length. Tags beyond roughly 60 characters get truncated in display, and long tags are more likely to be replaced entirely. There is no hard character limit, but the pixel width of the results column is real.
Keyword stuffing. Repeating the same term, or chaining variations separated by pipes, is the most reliable way to lose your tag. "Plumber London | Emergency Plumber London | Best Plumber London" will be rewritten almost every time.
Boilerplate repetition. A brand suffix on every page is fine. A whole template repeated across hundreds of pages, where only one word varies, invites replacement.
Mismatch with page content. If the tag promises something the body does not deliver, or the H1 describes the page more accurately, the H1 usually wins.
A fifth trigger shows up on ecommerce and directory sites specifically. When a page has many near-identical siblings — variations by colour, size or city — the tags become effectively interchangeable, and the system frequently substitutes a heading that carries the distinguishing detail instead. The fix is to put the distinguishing attribute early in the tag rather than at the end.
How to Check Which Pages Are Affected
Checking is faster than most people expect and does not require a paid tool. Search Console holds most of the answer.
Open the Performance report, filter to a page, and look at the queries with impressions. Then search two or three of those queries and compare the displayed title against your source. Doing this for your top 20 pages by clicks takes about 30 minutes and finds nearly every meaningful rewrite.
Crawlers make it faster at scale. Screaming Frog, Sitebulb and Ahrefs Site Audit will all export your specified tags, and comparing that export against what appears in results is a simple spreadsheet exercise.
The comparison that matters is between rewritten pages and their click-through rate:
| Signal | What it suggests | Action |
|---|---|---|
| Rewritten, CTR above average | Google's version is better | Adopt its wording |
| Rewritten, CTR below average | Both versions are weak | Rewrite from query intent |
| Not rewritten, CTR low | Title fine, intent mismatch | Check the page, not the tag |
| Rewritten to site name only | Tag seen as uninformative | Add the specific topic |
That first row is the one people resist. If the generated title is outperforming yours, the useful response is to learn from it rather than to fight it.
The Edits That Win the Tag Back
Fixes here are unusually reliable, because the triggers are mechanical.
Cut to 50 to 60 characters with the primary term near the front. Front-loading matters both for truncation and for the match assessment.
Use the term once. One clear mention beats three, and the variations you were chaining can live in the H1 and the opening paragraph instead.
Align the tag with the H1 without making them identical. When the two agree on subject and differ slightly in phrasing, the tag is far more likely to survive — a large gap between them is read as a signal that the tag is unreliable.
Make it describe the page rather than the site. "Pricing" is uninformative on its own and invites a rewrite to the brand name. "Pricing for Managed WordPress Hosting" is specific enough to keep.
Two edits are worth testing once the mechanical ones are done. Adding a number or a year works when the content genuinely supports it, and it is the most common source of a real click-through gain. Removing the brand suffix on informational pages frees 12 to 18 characters for something more useful, and brand recognition rarely drives the click on a how-to query anyway.
Measure both with a 28-day before-and-after on click-through rate at stable position. If position moved, the test is void.
When You Should Leave It Alone
Not every rewrite is a problem worth solving, and chasing all of them wastes time that belongs elsewhere.
If a page ranks for 40 different queries, a generated title that describes the page plainly can serve that whole spread better than a tag you wrote with one query in mind. Fighting that is a losing position, and the click-through data usually confirms it.
Check impact before editing. A rewrite on a page earning 12 clicks a month is not worth an afternoon. Sort by clicks and work down — in most audits, the top 10 pages hold most of the recoverable value.
One more case deserves patience rather than edits. Brand-name substitution on a homepage or a top-level category page is normal and usually harmless, because searchers arriving on those queries already know who you are. The click-through rate on branded terms is high regardless of what the tag says.
What Not to Do
Three common responses actively hurt. The first is adding data-nosnippet or trying to block the behaviour through directives — there is no supported control for title generation, and the workarounds people share tend to damage other things.
The second is rewriting the tag every two weeks. Title changes take time to be recrawled and reassessed, and rapid iteration destroys your ability to attribute any change in click-through rate. Change once, then wait a full 28 days before judging.
The third mistake is quieter and more expensive. Writing every tag to a rigid template — term, pipe, modifier, pipe, brand — recreates the boilerplate trigger at scale, and a library of 300 pages built that way tends to lose its tags all at once after a template change. Vary the structure across page types.
Templates are still fine for genuinely repetitive page sets like locations or products. The distinction is whether the varying part carries real information. "Emergency Plumber in Croydon" earns its template. "Services | Company Name | Official Site" does not.
FAQ
Does a rewritten title hurt my rankings?
No. Title generation affects what is displayed, not how the page is ranked. The indirect risk is click-through rate — a weaker displayed title earns fewer clicks at the same position, which is worth fixing for traffic reasons rather than ranking ones.
How long until my new title shows?
Usually 3 to 14 days, depending on how often the page is crawled. High-traffic pages update faster. If a month passes with no change, check the last crawl date in URL Inspection. If Google has recrawled the page since your edit, it still rates its own version as the better description.
Should the title and H1 be identical?
No, but they should agree. Keep the same core subject with different phrasing — the tag written for a searcher scanning results, the H1 written for a reader who already clicked. Large mismatches between them are one of the four common rewrite triggers.
The Practical Takeaway
Work by impact, not by count. Export your top 20 pages by clicks, compare the specified tag against what actually displays for their main queries, and fix only the ones where click-through sits below your site average. Cut each to 50 to 60 characters, use the primary term once and early, align it with the H1 without duplicating it, and make it describe the page rather than the brand. Then wait a full 28 days before evaluating whether it worked. Start with your five highest-traffic pages this week.
If you are standardizing titles across a large library, tools like FluxWriter can help apply a consistent pattern as posts are produced — but deciding which pages are worth the edit, and reading the click-through data that proves it, remains a manual judgement.