How to calculate the minimum sample size to analyze the influence of a factor on the CTR of advertising?

Question

Analytics

There is some hypothesis about the influence of a certain factor on the CTR of an ad.
It doesn't matter what the hypothesis is. Let's say demographics (age).
Let's say there are two samples - with selection by the criterion and without.
Let's say advertising is bought per click.
The expected difference in CTR is 1-5% of the average value (i.e. if without demographics it will be 0.5%, then with demographics, let's say 0.52%).
Should I have multiple pairs of samples to compare with each other?
Or is one pair enough - with and without?
What should be the sample size (and number of samples) for 95% confidence?
Something is completely messed up. It's not mine :)
UPD:CTR is click-through rate - click through rate. It is defined as the ratio of the number of clicks on a banner to the number of its impressions, measured as a percentage.

Answer 1

Two will be enough if the criterion is one (demography for example).
But let's say the difference is 0.5 and 0.52 at the level of error, either giant samples are needed here, and I don’t even know how to rely on them.