Click-Through Rate
The click-through rate measures how many people click on something that has been shown to them. It is one of the most important metrics in the online advertising business and indirectly controls what recommendation systems present to us in the first place.
The click-through rate is a simple calculation. You count how often an ad, a search result, or a video thumbnail appeared on screens. Then you count how often someone clicked or tapped on it. The number of clicks divided by the number of impressions gives the click-through rate, usually expressed as a percentage. If an ad appears 1000 times and is clicked 20 times, the rate is two percent. The English name has become established; in German it is also called Klickrate.
The currency of the ad market
Almost all major internet services are free and finance themselves through advertising. To do this, they need a number that shows whether an ad is actually working. The click-through rate is that number. It can be measured instantly, without a survey and without any waiting time.
For advertisers, real money depends on it. Many ads are paid for based on clicks, not impressions. An ad with a high click-through rate brings the platform more revenue per impression. That’s why advertising systems automatically favor the ads they expect to generate many clicks. A difference between 0.5 and 1.5 percent sounds tiny, but it triples the yield.
It’s important to distinguish this from the so-called conversion rate. This measures how many people who clicked actually go on to buy something or sign up. A click alone therefore says little about success. Sensational headlines often generate many clicks yet hardly any revenue. Experts call such titles clickbait.
How models predict clicks
Behind modern advertising systems lies a learning program. It estimates, for every possible ad, how likely a click is. To do this, it receives features such as time of day, device, past interests, and the content of the ad. From millions of past impressions, it has learned which combinations led to clicks.
This prediction then decides the auction. With every page load, several advertisers bid for the available slot. The platform multiplies the bid by the predicted click-through rate. Whoever achieves the highest value gets the slot. The whole process takes place in a few milliseconds while the page is loading.
To improve the predictions, variants are tested directly on the audience. In an A/B test, one half of users sees version A, the other version B. The version with the higher click-through rate wins. A well-known pitfall here is feedback loop: the system most often shows what it already considers successful. It then collects hardly any data on poorly rated ads and finds it difficult to correct itself.
Click-through rates in the feed and in quarterly reports
You encounter this metric daily without seeing it. The order of videos on YouTube or TikTok is also sorted by how often a thumbnail is clicked. Newsletter providers show their customers how many recipients opened a link. Search engines, too, evaluate which result users actually click on.
In business news, the click-through rate comes up around quarterly earnings. Companies like Meta or Alphabet regularly report on ad prices and click numbers. If the click-through rate rises, this is seen as a sign of better recommendation models. If it falls, stock prices often drop, because investors expect lower advertising revenue.
Typical values help put things in perspective. Classic banner ads often achieve only 0.05 to 0.1 percent. Ads above a search query, on the other hand, sometimes reach several percent, because interest is already present. A click-through rate without context therefore says almost nothing.
