There really isn’t one right way to do CRO. That said, you will want to follow certain guidelines to make sure that:
a) you are optimizing things that matter, and
b) you are able to see if your CRO efforts made a difference.
CRO starts from an issue or hypothesis. Perhaps you aren’t happy with your conversion rate, or get an idea about a change that could improve the conversion rate.
Before doing any optimization, find out what the current conversion rate is – otherwise you won’t know if your CRO efforts helped or not. After this, form a hypothesis about something that could improve the conversion rate.
Then, publish the change(s) you want to experiment. A/B testing is a commonly used method, in which you run two different variations of for instance a landing page (A- and B-version) in parallel. After the test, see which version performed better.
When running conversion rate experiments, it is usually a good idea to experiment with only one variable at a time. This ensures that you can attribute the changes in conversion rate to that variable.