Factors of selecting optimal number of personas

Our team has been working on the question of how many personas to create for a while now. It’s an elusive questions that keeps us searching for a final answer that may not be out there.

Regardless, we have been able to identify some factors relevant to this question. In this post, I’ll briefly go through these factors.

They are:

  • Technical optimality
  • User experience (UX) optimality
  • Task optimality

Technical optimality deals with the question, “does the data contain a lot or little variability?”. The more variability there is, the more personas are needed to adequately represent the data.

UX optimality deals with the question, “can the chosen persona medium deliver many or few personas in an efficient way?”. If personas are presented as A4 prints or posters, then it is difficult to have many personas because they would quickly become unmanageable to designers.

Taks optimality deals with the question, “are many or few personas needed for the job?”. Sometimes, practitioners want to focus on a specific persona (“give me one persona to design for”). Other times, they want to compare several personas (“give me all personas from Finland, so I can see their similarities and differences”).

Ultimately, there may not necessarily exist a magical number of personas that should always be created. Nonetheless, the question matters. If restricting the number of personas too low, we hamper the designer’s world-view, by excluding some groups that might be relevant. In contrast, if we present too many personas, we may alienate the designer, who feels they are lost in complexity.

People want simplicity, but at the same time they do no want to be fooled. To address this dilemma, we believe that tools are needed to help designers carve out a number that is just right for them – one that is faithful to the source data, manageable by using an interactive system, and works for the task at hand.

number of personas
Figure 1.¬†Description of the dilemma of having fewer or more personas. Dimensionality reduction and clustering algorithms tend to reduce the user datasets to a small number of personas, which may reduce designer’s ability to see variety.

 

1 thought on “Factors of selecting optimal number of personas”

  1. Pingback: What is "Persona Analytics"? – The Persona Blog

Comments are closed.