Did changing my page titles reduce my traffic?
Short answer
Compare CTR at held position on the changed pages against a control set you didn't touch, over the same window. Titles primarily affect click-through rate rather than ranking, so the signature of a bad title change is stable position with falling CTR on the changed pages only. Before concluding anything, check whether Google is even displaying your title, because it frequently rewrites them and your change may never have reached a user.
Check whether your title is being shown at all
This is the step almost everyone skips, and it invalidates the whole investigation when missed.
Google rewrites title links when it judges another version more useful, drawing from your H1, on-page text, or anchor text. If it rewrote yours, the title in your HTML isn't what searchers saw, and your change had no opportunity to affect anything.
Check by searching for the exact page, in an incognito window, and comparing the displayed title against your source. Do this for five or six of the changed pages. If most are being rewritten, your investigation ends here and the answer is no.
The comparison that actually proves it
A before-and-after on the changed pages alone cannot separate your titles from everything else that happened that month. You need a control.
Build the control set first. Pages you didn't touch, matched on intent bucket, traffic band, and page age. Twenty to fifty is enough.
Pull CTR at held position for both sets. Filter to queries where average position moved less than half a place, so you're measuring click behaviour rather than ranking.
Difference the two. If changed pages fell 18% on CTR and control pages fell 16%, your titles did almost nothing and something broader is happening. If changed fell 18% and control held flat, the titles are your cause.
| Changed pages | Control pages | Read |
|---|---|---|
| CTR down | CTR flat | Titles caused it |
| CTR down | CTR down similarly | Something sitewide, not titles |
| CTR flat, clicks down | Clicks down similarly | Demand or impressions, not titles |
| CTR up, clicks down | Any | Impressions fell, titles worked |
What usually goes wrong in a title rewrite
When titles genuinely cause a decline, it's rarely subtle wordsmithing. Four patterns account for most of it.
The primary keyword moved to the end or disappeared. Users scan the first few words, and a title that no longer visibly matches the query loses clicks.
Truncation. Longer titles get cut, and the cut often lands mid-phrase in a way that reads badly.
Brand added where it wasn't needed. Prepending your brand to every title consumes the visible space that used to carry the match.
The title stopped matching the query intent. A title optimised for a term the page ranks for secondarily rather than its primary driver.
Check your declining pages against those four before assuming the problem is stylistic.
The complication that's easy to misread
If titles changed at the same time an AI Overview started appearing on those queries, you'll see CTR fall on the changed pages and conclude your titles did it.
Check SERP composition for the affected queries alongside the title comparison. Pew Research found that when an AI summary appears, users click a source link in only about 1% of visits, which will dwarf any title effect.
Your control set protects you here too, provided the control queries didn't also gain Overviews. Check that they didn't, or the comparison is compromised.
If the titles did cause it
Revert the specific pages that lost CTR rather than the whole set. Some titles almost certainly improved, and a blanket rollback loses those gains along with the losses.
Then re-measure after a full four weeks. Title changes reach a stable CTR reading faster than ranking changes do, but a single week still swings on noise. See How long should I wait before deciding an SEO change failed?.
Run the test properly in Analyze AI
The reason this test usually isn't run properly is the control set. Building one manually, then pulling matched CTR data for both groups across two periods, is an afternoon of exports.
GSC Page-Keyword Breakdown returns every query driving clicks to a page, which lets you filter to held positions per page rather than site-wide. The Google Search node checks SERP composition on the affected queries so you can rule out an Overview appearing mid-test.
The test agent:
Start (schedule, weekly during the test) → Loop / For Each over the changed page set → GSC Page-Keyword Breakdown per page → Loop / For Each over the pre-registered control set → GSC Page-Keyword Breakdown per page → Google Search node on the top query per changed page (detect Overview or feature appearance) → Code (filter both sets to held positions, compute CTR delta per group, difference the two) → Conditional (if fewer than four weeks of post-change data, report as provisional rather than concluding) → Prompt LLM (state whether the difference is attributable to titles, and list the specific pages to revert) → Excel export → Send Email to the SEO lead.
The Conditional on elapsed time is worth copying. Title tests get called early more than almost any other change, usually in week two when the data still swings, and encoding the waiting period in the workflow removes the temptation.
The per-page revert list is the deliverable. Blanket reverts are what turn a mixed result into a net loss.
FAQ
Related answers
- Why is my organic traffic dropping while my rankings stay the same?
- How long should I wait before deciding an SEO change failed?
- Should I restore the old version of a page after traffic drops?
- What can I honestly claim from an SEO before-and-after analysis?
Want title tests run against a matched control automatically? Start a free Analyze AI trial and connect Search Console.