What is a good click-through rate in SEO?
Your CTR is some number and you have no idea whether it is good, bad, or exactly what it should be at your positions.
There is no good number, there is only your expected number
Click-through rate is almost entirely determined by position and query intent. A 2% CTR at position 9 is unremarkable. The same 2% at position 2 is a serious problem. Comparing your number to an industry average compares you to a different distribution of positions, which tells you nothing.
The comparison that works
Take your own curve, expected CTR at each position, then compare each page's actual CTR against what your other pages achieve at the same position. Pages below their expected value are being seen and not clicked, which is a title and description problem you can fix this afternoon. Pages at their expected value need a better position, which takes months.
Falling CTR is often growth
As a site ranks for more queries it ranks for more marginal ones, at worse positions, and the blended rate falls even while clicks rise. This site's CTR fell from 6.02% in April to 2.72% in July while clicks went from 4,409 to 16,410. The rate more than halved and the traffic nearly quadrupled, in the same months, for the same reason. Reading that fall as a problem would have meant fixing something that was working.
Good means better than your position predicts
There is no absolute threshold worth memorising. The only useful test is comparative: take the position, look up what your own non-brand curve returns there, and compare. A page at position 4 pulling 4% is underperforming badly. A page at position 14 pulling 4% has a title doing real work, and finding more like it is the cheapest win available to you.
Watch clicks, not the ratio
Click-through rate is a fraction, and fractions move for reasons that have nothing to do with performance. This site's overall CTR fell during a stretch when clicks were climbing, because impressions grew faster than clicks did, which is exactly what happens when you start ranking for more things. A falling ratio on rising clicks is expansion, not decline. Chase the numerator.
Measured, not asserted
Site-wide CTR here was 3.44% across 1,228,913 impressions, at an average position of 9.07. Brand queries, which click far above everything else, were 42.3% of clicks and had to be separated before any of the above made sense.
Free tool for this: Search Console Analyzer. No account, nothing uploaded.
Where this goes deeper
Every number on this page comes from one complete dataset: one product taken from zero to 100K+ users on search alone, with nothing spent on advertising. The full argument is Chapter 16 of the book. Five chapters are free to read.