When you have a lot of raw observations from research, but aren't clear what it all means.

Find themes in the research data by clustering findings. Then prioritise the themes.

1. Make sure you've got a big virtual whiteboard or a lot of wall space . Make sure you've got each research observation listed, one per real or virtual post-it note.
2. Start to put the post-its up, reading each one aloud as you do. This is important when you have a lot to sort through or a big group.
3. Each time you put a new post-its up ask yourself if it's related to a previous one that's already up there. If it is, place it very close to that existing note.
4. When you've run out of things to add (or before), stand back from the wall and look for clusters.
5. Using a way to differentiate them (use a symbol or a bigger post-it) write titles for these clusters and put them up over the clusters.
6. Review the outliers; they may belong to smaller clusters.
7. Your clusters are now your themes; you may want to prioritise these according to their importance.