One Fish, Two Fish, Red Fish, Blue Fish: Categorization in UX Design

Categories matter in data analysis and information design. It is tempting to take whatever categories the system already provides (emails opened, website visits, red and blue fish, etc.) and use them as is. That won’t necessarily give you the insights you need and you may overlook the most insightful way to optimally structure information. For this reason, qualitative research, like interviews, is often invaluable for the best categorizations.

Some additional things to know about quantification and categorization:

• A helpful approach to categorization is “identify first, quantify second.” Identify issues first, then measure how prevalent they are, when and with whom.
• The importance of categorization runs deep into any statistical analysis. When AI or statistical models get built, it is most often on the back of categorizations.
• Your models – whether qualitative or AI – encompass your understanding of the system and/or people within it, which are based on the categorization you have done. Therefore your decisions are also based on this categorization.
• The pragmatic way out of all these potential biases and flaws is to test your model and designs that arose from the insights it gave you.
• Lastly, categorization isn’t bad; it is necessary. It’s just important to be aware that you are making choices that are not handed down flawlessly from Mount Olympus – or an all-knowing AI database.