The essential characteristics of the users are collected (for example with the aid of data collection via surveys or evaluation of user data). The user groups are characterized in detail by extracting example users from these existing data. Out of this, personas are developed and brought to life. This already happens during the project delimitation. The example users get hypothetical (but obvious) personal attributes such as name, looks, job, education, interests, skills, experiences, etc. At the end, you have to make a comparison to see if you have reached your goals.
There are a lot of discussions going on if personas are a helpful method or not. Some experts claim Jobs-to-be-Done (focussing of the real needs and jobs of people instead of there demographic information) is much more valuable and relevant (see links) than discussing personal traits.
The Nielsen Norman Group recommends to combine Personas and Jobs to be Done, as "the usability of any given design can only be assessed relative to two variables: who are the users and what do they need to do? That’s why it’s critical for the validity of a usability study to recruit representative test users and give them representative tasks to perform."
We collected some diverging opinions and controversal articles on the topic (see links). It always depends. We recommend to make your own picture. :-)
Anything to improve with the method description? Just send an e-mail with your suggestion. Thank you!