Search results and measurement
CTR: calculate and interpret click-through rate
CTR, or click-through rate, is the number of clicks divided by the corresponding number of impressions, usually expressed as a percentage. Both figures must refer to the same system, surface, period and filters. CTR is neither a purchase rate nor a percentage of unique people.
Also known as: click-through rate
At a glance
- Calculate clicks divided by matching impressions, multiplied by 100.
- Combine groups using their total clicks and impressions, not a simple average of percentages.
- There is no universal good CTR independent of channel and search context.
On this page

How is CTR calculated?
CTR = clicks ÷ impressions × 100. For example, sixty clicks from two thousand impressions produce a CTR of 3%. This is an illustrative calculation, not an industry benchmark.
Search Console and Google Ads both use this ratio, but cover different events and surfaces. Do not divide organic clicks by advertising impressions.
With zero impressions, the ratio is undefined. “No calculable data” is more accurate than inserting 0%. Zero CTR is valid when impressions occurred but no corresponding clicks did.
What must match in the numerator and denominator?
Impression rules depend on the system and result type. Google's counting guidance describes those differences.
Keep the period, search type, country, device and page or query scope consistent. Hungarian mobile web-search results last month are not the same group as all countries and devices.
Clicks are not necessarily distinct people. CTR summarizes event counts rather than the proportion of individuals interested in a product.
Why can averaging percentages be wrong?
Imagine two non-overlapping query groups for a Budapest language school, covering all observations in the example.
| Group | Impressions | Clicks | CTR |
|---|---|---|---|
| Searches naming the school | 100 | 20 | 20% |
| General course searches | 1,900 | 38 | 2% |
| Total | 2,000 | 58 | 2.9% |
A simple average of 20% and 2% gives 11%, which is incorrect because one group has nineteen times as many impressions. The correct combined rate is 58 ÷ 2,000 × 100 = 2.9%.
These fictional figures demonstrate why counts matter. If the group mix changes next month, overall CTR can change even when each group's rate stays constant.
What counts as a good CTR?
The percentage needs context. A branded search differs from broad research, and advertising differs from organic search. Your own comparable queries and pages are often a more useful reference than a universal target.
Inspect the SERP layout and competing elements. Those observations suggest explanations without proving a single cause.
Small samples fluctuate sharply. Two clicks from twenty impressions is 10%; three clicks is 15%. A five-percentage-point difference can represent just one additional event.
What can a change suggest?
Check whether the query mix changed, whether the title tag accurately describes the page, and whether the meta description communicates a relevant distinction. These are hypotheses to investigate, not causes revealed automatically by the metric.
Record changes and dates. If title, article and offer all change together, attributing the outcome to one element becomes difficult. Compare clicks, impressions and group composition alongside the percentage.
What does CTR not reveal?
It does not show whether readers purchased, inquired or found the answer. A misleading promise can increase clicks while causing disappointment. Measure meaningful outcomes separately.
Conversion rate, engagement rate and bounce rate use different events and denominators. A low site CTR also does not prove zero-click search: users may have clicked another result.