Interviews produce noise before they produce insight. The Kraft Heinz "Discovery of Pain" boards and the EU Connected Dispenser bubble maps are both the same synthesis move: cluster verbatim pain points by theme, then size each cluster by how often it actually came up, not by which quote is most quotable.
The clustering matters less than the discipline of keeping quotes attached to their theme instead of summarizing them away too early. "Downtime management," "labor shortages," and "data analytics" only became distinct, prioritizable themes in the Kraft Heinz work because each one still carried three or four direct quotes underneath it when it went in front of stakeholders.
A synthesis pass is ready to present when:
- Every theme is still traceable back to specific interview quotes, not just a label someone remembers.
- Themes are sized by frequency across interviews, not by the loudest single voice.
- The synthesis produces a ranked shortlist, not just an interesting collage - someone has to be able to point at what to build first.