A useful starting target
For a non-branded Google web search, a CTR around 20% at position 1, 6% at position 3, or 1% at position 10 is a reasonable starting reference. At position 8, 2% is a healthy result. At position 1, the same 2% deserves a closer look. Apply these to a specific query or a small, similar group, rather than your whole site.
Use the following bands to decide where to spend time. They are Numic’s editorial review thresholds, informed by the research below; they are not Google grades or measured percentiles. The starting case is a non-branded query where someone has a reason to visit a website. Keep the country, device, and reporting period consistent.
| Approximate position | Healthy starting range | Investigate below |
|---|---|---|
| 1 | 15–25% CTR | 10% |
| 2 | 8–12% | 5% |
| 3 | 4–8% | 2% |
| 4–5 | 2–5% | 1% |
| 6–10 | 0.5–2% | 0.3% |
Treat results above the healthy range as a positive sign. For values between the two columns, compare with your own similar queries before editing. Below the review threshold, inspect the result and the search intent. A low CTR identifies something to investigate; it does not tell you which change will help.
What the published data says
Advanced Web Ranking’s July 2026 US data, across devices, reports non-branded CTR of 19.38% at position 1, 6.02% at position 3, and 0.81% at position 10. These are sample averages, not minimum acceptable results. See the branded/unbranded breakdown in its organic CTR study.
An AI Overview changes the comparison. In that study’s separate US desktop sample, position 3 averaged 1.70% with an AI Overview, versus 3.89% across all result layouts. The all-layout figure includes AI Overviews; it is not a clean “without AI” control. These observations do not prove that adding an Overview caused a particular page’s decline.
Use the first table for an initial review, then use a matching country/device/layout segment if available. The numbers here are a dated reference, not a live benchmark feed. AWR’s monthly study tool lets you inspect other segments and months.
Five examples with a verdict
These are illustrative reports and editorial judgments, not customer results. Each uses one query and page, one device and country, and a 28-day period.
1. A buying guide at position 3 with 6% CTR: good
A guide to choosing a compact coffee grinder gets 10,000 impressions and 600 clicks. The results mainly link to reviews and product comparisons. That is a healthy result. Leave the title alone unless it is inaccurate; improving the guide or moving up a position is a more sensible next experiment than chasing an arbitrary 10% CTR.
2. A useful tutorial at position 1 with 2% CTR: weak
A tutorial on repairing a leaking tap gets 10,000 impressions and 200 clicks, with no prominent answer or video replacing the need to visit. Investigate. Does its title describe the right type of tap? Does the snippet promise the steps the searcher needs? Compare the displayed result with nearby results before rewriting the page.
3. A definition at position 3 with 2% CTR and an AI Overview: reasonable
On US desktop, a page explaining “what is a canonical URL” gets 200 clicks from 10,000 impressions. The Overview answers the basic question before the links. We would not label 2% a title failure here. Check whether the page offers a useful next step—examples, troubleshooting, or a working tool—and whether those are visible in its search result.
4. Your exact brand at position 1 with 8% CTR: investigate
People searching a distinctive business name are often trying to reach that business. Eighty clicks from 1,000 impressions is low enough to inspect. Check for competing ads, a different business with the same name, or a result pointing to the wrong page. Our working target for a clear “take me to this website” query is at least 30%. That is an editorial target for navigation intent, not an average for every query that happens to contain a brand name.
5. A comparison page at position 8 with 2% CTR: good for its rank
At 10,000 impressions and 200 clicks, the page is already attracting visitors despite being farther down the results. Work on whether it deserves to rank higher: missing comparison criteria, outdated information, and relevant internal links. The title is not the first suspect.
Choose the next step by the size of the opportunity
Suppose a relevant page at position 3 has 20,000 impressions and 400 clicks: 2% CTR. If similar queries on your site reach 4%, that is a useful test target. At the same impression count, 4% would mean 800 clicks, or 400 more. A page with only 200 impressions would gain four clicks from the same two-point increase. Start with the larger opportunity if the visits are equally valuable.
This is a scenario, not a prediction that a title edit will double traffic. The CTR comparison calculator can do that comparison and keep percentage points separate from relative change. For separating brand variants first, use the regex builder.
If CTR is healthy but the ranking is weak, use the position guide to decide whether ranking work is the better investment. If the whole site’s CTR fell, inspect its query mix before applying these thresholds to every page.
How much data is enough to judge?
Five clicks from 100 impressions is 5%; two more clicks would make it 7%. Do not rewrite a page over that movement. As a practical screening rule, start with 28 complete days and around 1,000 impressions in the group you are judging. Extend to three months for a smaller site, checking that position and demand have not changed substantially. This is a review rule, not a statistical confidence threshold.
An 8% CTR for the whole site does not receive a useful good/bad grade. Split it into branded and non-branded queries, then into position groups. If 8% belongs to non-branded position-3 searches, it is strong. If it belongs to your distinctive brand at position 1, it deserves attention. That split tells you what to do.