Affinity maps are momentary by nature and are a great device for externalising information, making them an excellent tool for problem-solving in agile teams.
However, they have their setbacks; they don't support remote collaboration, quick recall of information or large and dynamic data sets. When analysing user research data, our team found that we needed the ability to quickly identify relationships in qualitative data without these restraints that exist with paper documentation. Applied unconventionally, we found that Trello gave us the ability to do just that.
We first developed this technique when we had the challenge of analysing an extensive data set of over 15 scribed lab testing interviews. The process gives us the ability to group information semantically and filter it further for quantitative analysis. I've taken a small data set from a Website Usability Testing session to demonstrate the process below.
This is the basic process that can be adapted to suit many different use cases:
The initial state of the board is simple. It consists of two lanes; one titled 'Resources' and the other 'Feed'. The resources lane acts as a reference to sources relevant to the research. Resources include the research goal, interview guides, participant descriptions and personas. The 'feed' lane houses all unfiltered cards until they relocate to a group on the board.
These attributes are necessary for filtering and quantifying the cards once they have been sorted into groups. Enable the 'Custom Fields' power up and add fields for the attributes that you would like to use to identify and filter the cards. Tip: I estimate the effort required to input the data into the card fields and measure it against the value it intends to provide.
For this challenge, I created four attributes for the following reasons:
Type or copy each piece of important information into a card, then define the attributes. This process can take some time if you haven't received your research data in an organised fashion. We highlighted meaningful quotes when initially recording the research to make this process easier. If your notes are on sticky notes, the Post-it App scans your notes and directly transfers the text on each post-it to a Trello card - neat handwriting is essential.
Now the fun begins! Unpack the problem or opportunity by organising the data into meaningful groups. Work your way down the feed and group cards with similar themes and or patterns in the same lanes. Name each vertical data group, defining the essence of what the theme is.
Use the search function to filter by card attributes to visualise the relationships between the data in different ways or to quantify groups. In the example shown, I have filtered the lanes to show cards that are related to Prospective Customers and Services on the website. From here, I can calculate that 88% of comments made by this user group are related to services, with the majority linking to the user's inability to relate their problem to our service descriptions. This example is quite simple, and an experienced researcher may be able to identify it without this tool. However, as data intensifies problems become harder to refine, making this tool an ideal candidate for more complex scenarios.
Keep exploring and refining each data group until you unearth coherent and distinctive data groups. I often create off-spring boards to analyse the themes further.
A pitfall of this process is the inevitable requirement of screen time and the reduced creativity that accompanies it. I still like to apply traditional affinity mapping techniques if the process permits.
We'd love to hear your feedback and suggestions!