SEO click-through rate: analyse results and test titles
Click-through rate, or CTR, measures the share of search impressions that lead to clicks on your page. To improve it, first understand the searches involved, then check whether the title and content offer a suitable answer. A higher rate is useful when visitors find what they need and can take a relevant next step.
This guide provides a method for planning your next SEO tests, including your 2027 plan. It separates available measurements from assumptions about search results and AI-generated answers.
Calculate CTR and interpret changes
CTR = clicks ÷ impressions × 100. In a fictional example, 50 clicks from 1,000 impressions produce a CTR of 5%. The next week, 80 clicks from 2,000 impressions produce 4%. The rate falls while visits increase. Read all three figures together.
There is no universal “good CTR” for each position. Searches for your company name, general questions and product searches offer different opportunities to click. The results displayed, country and device also affect the situation. An external study cannot replace your own data.
1. Select pages and searches to investigate
In Google Search Console’s performance report, examine clicks, impressions, CTR and average position. Filter a page and then its queries. Separate branded searches from other searches and compare the same countries, devices and search types.
Identify relevant queries for which a page receives impressions but few clicks. Then check its position and how it answers the need. A page appearing for searches outside your offer requires a different decision from a service page with a strong position.
Record a baseline with the period, filters and values. Compare periods of similar length; note seasonality, campaigns and position changes. Average position combines several situations and does not mean every visitor saw the same result.
2. Make the title’s promise specific
Write a unique title that identifies the page’s subject and contribution. “Technical SEO audit: checks and priorities” explains more than “The ultimate SEO secrets”. Align the title with the actual content and visible page heading.
Google can construct the displayed title from several elements. Display width varies: a 50- or 60-character rule cannot guarantee an uncut title. Put useful information first and avoid repetition. See Google’s title guidance.
Add a year when the subject depends on that period and you have updated the substance. Do not replace “2025” with “2027” on an unchanged guide. Historical dates, such as when a rule took effect, retain their meaning.
3. Explain what visitors will find
Write a specific meta description covering the answer, scope and information that helps readers choose. For a service page, describe the work involved. For a guide, identify the decisions or checks explained. Avoid promising a free service, price or result that the page does not substantiate.
Google often generates the snippet from page content and may show a different passage for each query. The meta description is a suggestion; its display is not guaranteed. Google sets no maximum character length and may truncate the snippet to suit the device. Snippet documentation.
4. Check the answer and structured data
If a specific query reveals an important need, answer it on the appropriate page. Long-tail phrases help you understand that need; creating a page for every close variation can fragment your answers. Our SEO strategy guide explains how to organise content.
Use structured data suited to the content visitors can see, following the properties and conditions for the relevant type. It guarantees neither a rich result nor a higher CTR. Check errors and eligibility before deployment.
Google removed FAQ rich results in May 2026. Keep FAQs that help readers; stop presenting their markup as a way to obtain those results. Official Search Central update log.
5. Test a change and track its consequences
Choose a hypothesis: “Visitors do not understand that this page explains ongoing costs.” Revise the title to make that point clear after checking the content. Record the date and allow Google to crawl the new version before interpreting data.
Then compare your selected segments. A simple before-and-after comparison is not a controlled A/B test: seasonality, competitors and ranking changes may explain the difference. If you test groups of pages, check their comparability and document limitations. Do not draw conclusions from a handful of impressions.
Track relevant enquiries, orders or the actions you have defined. A title that attracts more clicks but disappoints visitors may reduce traffic quality. Our marketing measurement guide helps connect acquisition with results.
Prepare your 2027 measurement plan for AI results
Google announced and rolled out dedicated impression views for its AI features in Search Console during 2026. Use this information to examine visibility. These views do not, on their own, provide CTR or sales attribution for those answers. Report announcement and scope.
For your 2027 plan, record separately the impressions, clicks and commercial actions you can measure. Examine search-result changes for priority queries and reassess your hypotheses. Avoid presenting a fall in overall CTR as proof of an AI effect without a suitable comparison.
Frequently asked questions
Should we change every page with a CTR below 2.5%?
No universal threshold justifies that decision. Examine query relevance, volume, position and resulting actions before selecting pages.
Does a new title guarantee more clicks?
Its effect depends on the display, competing results and expected answer. Google may also select another title. Test a hypothesis and check what appears.
Where should we start?
Choose an important page and a relevant query group. Record your data, check the content and then improve one specific element. If you need support, our search engine optimisation service covers analysis, content and technical checks.