Why a CTR can only be compared with itself
CTR depends first on the delivery context, well before the quality of the ad. A search ad that answers an explicit intent will inevitably get a click-through rate on a different scale from a display banner shown to someone who was not looking for anything.
Comparing the CTR of two different networks therefore makes no sense. Even within a single platform, the format, the placement, and the type of audience change the scale of values.
The only useful comparison is over time and like for like: your CTR today against your CTR last month, on the same campaign and the same format.
When a high CTR is a bad sign
A high click-through rate combined with a low conversion rate signals a broken promise: the ad attracts clicks it does not convert. This is common with overly broad hooks, or with promotions advertised without their conditions.
Those clicks cost money and also pollute the learning of bidding algorithms, which receive signals of interest that are not followed by a conversion.
CTR is therefore an intermediate metric: useful for diagnosis, never enough to make a decision. Budget allocation decisions are made on cost per acquisition or profitability, not on click-through rate.
Frequently asked questions
How do you calculate CTR?
Divide the number of clicks by the number of impressions and multiply by one hundred. 300 clicks on 20,000 impressions give a CTR of 1.5%.
What is a good CTR?
It depends entirely on the network and the format: search and display values are not on the same scale. The only reliable benchmark is your own history, like for like.
Is a high CTR always good?
No. A high CTR combined with a low conversion rate signals an ad that promises more than the page delivers. Those clicks cost money and degrade the learning of automated bidding.
Does CTR affect bidding costs?
Yes, indirectly: platforms favor ads that engage, which generally translates into cheaper access to delivery. It is a consequence of relevance, not a lever to optimize for its own sake.