What is organic CTR, and is a low one actually a problem?
Organic CTR is the share of people who see your page in Google's unpaid results and click it. Search Console defines CTR as "the click count divided by the impression count." A low one isn't a problem on its own, because both halves depend on things a title can't change: where the page sits, what the searcher wanted, and whether Google answered them before they needed to click. So before I touch a title, I work out which of those is driving the number. When the cause is Google answering on the results page, the job changes from chasing clicks to being the page Google quotes when it answers the question itself, and that's a different piece of work.
The impression side is looser than most people think. Search Console counts an impression whenever a link appears on the current page of results, whether or not it's scrolled into view. A page at position nine collects an impression from every searcher who stopped reading at the first result. Those impressions are real, but plenty of them were never seen.
Brand searches pull the other way. On a villa rental site with 117,541 impressions at a 0.88% click rate, the query 'areca homes' clicks at 52.43% from position 1.4, while the site as a whole averages 8.9. Someone typing a company's name wants that company. Leave those queries in and every other page looks worse against the site average than it really is.
AI answers are already inside these numbers, too. Google reports sites that appear in AI Overviews and AI Mode in the Performance report under the Web search type, so there's no separate line to subtract. A low rate is a question, not a verdict.
What's a good organic CTR for each position?
Whatever the site itself earns at that position, not a figure from a chart in someone else's study. Published click curves come from other sites and other results pages, which won't look like yours, so I build the benchmark from the site itself. I export the queries from the Performance report, take the brand searches out, group what's left by average position and note the typical click rate in each group. A page then gets judged against its own band.
Average position is an average, so I read it per query rather than per page. A page that sits at three for one query and at forty for another shows a middling figure that describes neither.
| Position band | What I expect to see | What I check before the title |
|---|---|---|
| 1 to 3 | The site's highest click rates, unless something sits above the result | Whether ads, a map pack or an AI Overview push the link down the screen |
| 4 to 6 | Solid rates on commercial queries, softer ones on questions | Whether the title is cut off or reads like every other result |
| 7 to 10 | Low rates, and a lot of impressions nobody scrolled to | Whether the query is informational before I touch anything |
| 11 and beyond | Close to zero clicks, which is normal on page two | Rankings first, because a title can't fix a page nobody sees |
One site shows the effect clearly. Areca Homes' phone searches average position 7.9 and click at 1.03%, while desktop sits at 10.6 and clicks at 0.55%. Same site, same titles, and the worse average position comes with roughly half the rate. Device isn't the only difference between those two groups, but it's why I don't call a title weak until it's underperforming other pages at the same position.
Which queries will never click, and why?
The ones the results page already answers. A zero-click search is one where the searcher gets what they need from Google itself: a definition, opening hours, a date, a quick fact, or an AI Overview that sums the answer up. The link still earns an impression. It just doesn't need a visit.
Travel questions are full of them. On Areca Homes, 'blue hole jamaica' drew 7,811 impressions at position 9.0 and clicked at 0.24%, because the basics a traveller wants show up in the results before anyone needs a guide. Definitions behave the same way. Bright Pathways ABA's guide to the four functions of behavior drew 45,411 impressions and 111 clicks at position 9.4, the clearest case I've got.
I don't fight these with bait titles. A title that promises more than the page gives might win a click, but the searcher leaves again, and the next person still gets the answer from the results page. I either leave a page like this alone or work on making it the clearest answer on the subject, which is the part a site controls.
What I won't do is count these impressions as failure. They put a business in front of people early, before they're ready to buy; they just don't show up as clicks.
Google offers nosnippet and max-snippet controls to limit what it shows from a page. I don't use them as a click trick. They hide the text a searcher would've judged the page on, and a result with less to read isn't a stronger one.
Rather have me do this? One person, whole job.
Book a call on UpworkHow I find the pages worth rewriting
This is the order I work in on a site I haven't seen before. It turns a long export into a short list of pages where a better title and snippet have something to win.
- Step 01
Export three months of queries and pages
From the Performance report, Web search type, I pull queries and pages for the last three months. A shorter window is too noisy on a small site, and a longer one mixes in pages that have since moved.
- Step 02
Take the brand searches out
Anything with the business name goes in its own pile. Those queries click well because the searcher already chose the business, and they'd hide the pages that really need work.
- Step 03
Tag every query by intent
Informational queries want an answer. Commercial ones want a provider, a price or a product. On Areca Homes, the property management page clicks at 2.01% from 6.6, more than double the site's rate, because the owner searching for it wants someone to run their villa.
- Step 04
Compare each page with its band
Each page gets checked against the typical rate for its position band. A commercial page clicking well below its band is a candidate. An informational page below its band usually isn't, for the reasons in the section above.
- Step 05
Read the live results page
I search the query myself and look at what sits above the link: ads, a map pack, an AI Overview, videos or a block of related questions. I also check which title Google's actually showing, because it isn't always the one in the page's title element.
- Step 06
Check that one page owns the query
When two pages trade places on the same search, each gets a share of the impressions and neither earns the clicks it should. A cannibalization check on the same export finds those pairs. Then I settle which page owns each query with a merge or a clearer split, which usually does more than a new title would.
- Step 07
Sort by what's left to win
Impressions times the gap to the band's typical rate gives a rough count of missing clicks. Commercial pages go to the top, and I start with the three or four at the head of the list, not all of them at once.
Don't rewrite titles on pages sitting past position ten. Hardly anyone reads page two, so a better title can't lift clicks there. Those pages need ranking work first, and the click rate follows the position.
How I improve organic CTR with titles, snippets and rich results
Only the parts a searcher reads before choosing. Google builds the title link from several sources, among them the title element, the page's main visible title, its headings, og:title and the text of links pointing at it, so those need to agree. The snippet mostly comes from the page content, and Google sometimes uses the meta description instead when it gives searchers a more accurate summary.
- i.
Lead with the words they typed
The query's own wording goes near the front of the title, once. Google's documentation warns against repeating the same words, and a title that says the thing plainly beats one that says it three ways.
- ii.
Say what the page adds
If the results page already covers the basics, the title has to offer what it can't. For a travel guide on a villa site, that's a villa a short drive from the place the searcher's reading about.
- iii.
Match the title to the H1
When Google decides the title element doesn't describe the page well, it may build the title link from the main heading or other text on the page. I make the two agree, so the words I chose are the words most likely to show.
- iv.
Write the description with specifics
Google's advice is a unique description per page with real details in it, not a string of keywords. I put in what the title can't carry: the area covered, the format, or who the page is for.
- v.
Keep every title distinct
Boilerplate repeated across pages makes results hard to tell apart. Each page gets its own title and its own description, and the site name goes at the end, if it goes in at all.
- vi.
Earn a richer result where the page qualifies
Structured data can make a page eligible for a rich result, and Google's own case studies include Rotten Tomatoes measuring a 25% higher click-through rate on pages with it. Google also says it doesn't guarantee the markup will show, so I only add types the visible page genuinely supports. Featured snippets can't be requested at all: asked how to mark a page as one, Google's answer is "You can't."
- vii.
Change a few pages, then wait
Google has to recrawl and reprocess a page before a new title shows, which it says may take a few days to a few weeks. I note the date of each change, then compare 28 days after with the 28 days before for the same queries.
If the click rate rises and the position holds, the title did its job. If the position moved as well, I can't credit the title alone, and I'll say so in the report rather than claim the win.
If you'd rather one person worked through the export, sorted the queries and rewrote the titles that can actually win clicks, that's work I do remotely on WordPress sites. I'll also tell you plainly which pages aren't worth touching.
Book a call on Upwork
