How to write a title tag that wins the click

The words that win the click are the ones people already type to find you, and they are sitting in your Google Search Console query view. To write a title tag that earns more of the clicks you already rank for, pull the queries the page shows up for, find the ones your current title doesn't contain, write the title (and the H1) around them, then measure the click rate afterward. This guide walks that query-view-to-title workflow step by step, plus the honest part no title checklist mentions: Google rewrites most of the titles it's handed, so the real goal is to influence the rewrite, not to pretend you control it.

This is the how-to companion to spotting the opportunity in the first place. If you haven't yet separated the pages worth a rewrite from the ones losing the click to something a rewrite can't fix, start with low CTR at a good position: title or AI Overview, then come back here to do the actual writing.

TL;DR:

• The best title words are the queries you already rank for. Your query view lists them, and the ones your current title omits are the phrase gap to close.
• Read those queries at the page level for volume, and use the query list only for the words, because Search Console hides rare queries and the query totals under-count.
• Put the searcher's words near the front, keep the title around 51 to 60 characters, and match your H1 to it. In the largest public study, ideal-length titles were rewritten least, and matching the H1 cut rewrites sharply.
• Google still rewrote 61.6% of titles in that study, so treat yours as a strong influence, not a guarantee, and confirm it worked by reading the page-level click rate after you ship.

Where do the words that win the click come from?

They come from your own search data, not a keyword tool. Every page that ranks already shows up for a set of real queries, and Search Console lists every one of them. The words in those queries, especially the high-impression ones your current title never says, are the exact phrases a better title should use. You are not guessing what searchers want. You are reading what they already typed.

This is the difference between writing a title and writing the right title. A generic checklist tells you to "use your keyword," but your page rarely ranks for one keyword. It ranks for a spread of real searches, and the page-level versus query-level split is what lets you see that spread honestly. The gap between the words in those searches and the words in your title is the opportunity, and it is specific to this page.

How do you read the phrase gap in your Search Console data?

Pull the list of queries the page ranks for, sort it by impressions, and scan for searches whose words your current title leaves out. A page shown thousands of times for a phrase whose title never contains that phrase is leaving the scan on the table: a reader deciding between ten blue links picks the one that visibly says the thing they searched for.

One honest limit governs how you read this. Take the volume, the impressions and the click rate, from the page level, and use the query list only for the words. The per-query counts under-report, because Search Console hides rare queries while still counting their impressions in the page total, so the query view is a reliable guide to what you rank for and an unreliable guide to how much. For writing a title that is exactly the right job: you want the words, and the words are all there.

A page ranking for searches it was never built to answer is a different problem, and no title rewrite fixes it. If the query list is full of the wrong kind of search, you are looking at an intent mismatch, not a title gap.

How do you write the title once you have the words?

Put the searcher's words near the front, be specific over clever, and keep the whole thing around 51 to 60 characters. That length matters for a concrete reason: in a study of 80,959 titles across 2,370 sites, Zyppy found titles in the 51-to-60-character band were rewritten least often, between 39% and 42% of the time, while shorter and longer titles were rewritten far more. Front-loading the words also survives truncation, because Google cuts desktop titles at roughly 600 pixels, near 60 characters, and a keyword that falls off the end helps no one.

The move most people skip is the H1. Match your on-page H1 to the new title, because the same study found that matching the H1 to the title "typically dropped the degree of rewriting across the board, often dramatically." That is not a coincidence: Google builds the title it shows partly from your page's own headings, so a title and an H1 that agree give it fewer reasons to overrule you. Write both from the same query words and you line up the two signals Google leans on.

The description is the argument for the click

The title earns the scan, and the meta description makes the case. Write it as the reason to click, not a restatement of the title: name what the reader gets, in their words. Google rewrites descriptions more often than not, but when it does show yours, a description that gives a reason to click still beats a summary that gives none. Keep the primary phrase in it, because Google bolds the words that match the search.

Does Google even use the title you write?

Often not verbatim. In that same study, Google rewrote 61.6% of all titles, and even the best-behaved, ideal-length ones about 40% of the time. Google's own documentation is blunt about why: the title link is "completely automated" and drawn from your <title>, your headings, other prominent on-page text, and even the anchor text of links pointing to the page. You influence it. You do not control it.

The good news is that everything in the section above is how you influence it in your favor. A descriptive, specific title that matches the query, sits at a sensible length, and agrees with the H1 is exactly what Google's generator is built to keep. Write a vague or stuffed title and you hand it a reason to replace you with something assembled from your page. Write the query-matched one and you are giving it the version you want it to pick. Treat the rewrite as a bias you are nudging, not a switch you are flipping, and the odds move with you.

What if a better title still doesn't win the click?

Then the title was probably never the problem. A page holding a strong position with high impressions and a click rate near zero is usually losing the click to an AI Overview answering the query above it, and no title, however sharp, gets that click back. Rewriting it is an afternoon spent where it cannot help. The title-or-AI-Overview read is worth running before you write, precisely so you spend the effort on the pages a rewrite can actually move.

The tell is the shape of the numbers: a good position and healthy impressions with a click rate that is merely below your own baseline is a title job, while a click rate near zero at a strong position is a SERP that changed above you. One wants the words in this guide. The other wants your page to become the thing that answer cites, which is a different move entirely.

How do you know the new title worked?

Treat the edit as a hypothesis and read the page's click rate over the next few comparable weeks, on page-level data so the hidden-query gap doesn't distort the before-and-after. Change one page's title at a time and note the date, or you won't know which edit moved the number. Give it a couple of weeks, because rankings and the click rate both take time to settle, and the newest Search Console data is still preliminary.

Two honest limits keep the read grounded. Position is a blended average across every query and searcher who saw the page, so a page "at position 6" was second for some searches and fifteenth for others; treat it as a rough band, not a coordinate. And the whole exercise stops at the click. A better title earns the visit, and whether that visit converts is a separate measurement in your analytics or your product, not in your search data.

Doing this by hand for every under-earning page is the tedious part: pull each page's queries, find the words the title omits, check that an AI Overview isn't the real cause, write the title and the H1 together, then remember to snapshot a baseline so you can tell later whether it landed. That is the loop QueryScope runs in your terminal. It reads your real Search Console data, surfaces the pages ranking well but under-earning the click, and keeps the daily history that tells you whether a title edit actually won the clicks back. If you want the one-line definition behind click rate or position first, the Search Console glossary covers each metric with the caveat that comes with it.

Sources

• Zyppy (Cyrus Shepard), Google Rewrites 61% of Page Title Tags (study of 80,959 title tags across 2,370 sites: Google rewrote 61.6% of titles; the 51-to-60-character band was rewritten least, 39% to 42%; matching the H1 to the title "typically dropped the degree of rewriting across the board, often dramatically").
• Google Search Central, Control your title links (title link generation is "completely automated" and draws from the <title> element, headings, prominent on-page text, and anchor text of links to the page; "write descriptive and concise text for your <title> elements").
• Google Search Central, A deep dive into Search Console performance data filtering and limits (rare queries are dropped from the query table while their impressions stay counted in the page total, so page-level data is the honest source for volume and the query view is for the words, not the totals).