It depends what type of testing you’re doing. There are testers who are meant to use the software like a typical customer, give feedback on what the user experience is like, and maybe see if they can break the system by typing in incorrect commands, or doing something else unexpected.

But the more serious testers for most software will need to do a lot of coding to write up test suites. Essentially, once you’ve written some code which interfaces with the software, you can automate the process of trying different inputs and returning metrics - not only whether the code crashes, but also how fast it is on different types of input, and where the bottlenecks are.

I’d say the latter type of tester is more common, since for every finished product which needs to be tested by using it, there are many different internal modules which need to be tested by actually writing code.