From 20 Interviews to One Decision
Synthesis is where research lives or dies. A repeatable method for turning messy interview notes into a single, defensible product decision.
The synthesis bottleneck
Most research does not die in the field. It dies in synthesis. You run 20 interviews, fill a spreadsheet with quotes, tag everything with themes, and then... nothing. The findings sit in a deck that no one reads. The decision is made by intuition, and the research was a checkbox.
The problem is not that synthesis is hard. It is that teams treat synthesis as a sorting exercise rather than a decision exercise. Tagging quotes by theme is not synthesis. Synthesis is the act of turning patterns into a recommendation that a product team can act on next Monday.
A repeatable method
Start with the decision you are trying to make. Before you run a single interview, write down the question the research is supposed to answer. Not "understand the user" — that is not a question. "Should we build a calendar view or a list view for the scheduling feature?" That is a question. Every interview should be designed to gather evidence for or against that question.
After the interviews, do not start with themes. Start with the decision. Write the two or three options you are considering at the top of a whiteboard. Then go through your notes and pull every piece of evidence that supports or undermines each option. You are not looking for themes. You are looking for a verdict.
The third step is the confidence check. For each option, rate your confidence: high, medium, or low. If your confidence is low across the board, you do not have enough evidence. Run more research. If one option is high and the others are low, you have your answer. If they are all medium, look for tie-breakers: cost, effort, risk.
Writing the synthesis
The output of synthesis is not a deck. It is a one-page document with three sections: the decision, the evidence, and the confidence. The decision is one sentence. The evidence is three to five bullet points, each with a quote or data point. The confidence is a single word with a one-sentence explanation.
This format forces discipline. If you cannot write the decision in one sentence, you have not synthesized. If you need more than five pieces of evidence, you are probably cherry-picking. If your confidence is "medium" with no explanation, you are hedging. The one-pager is not a limitation; it is a quality bar.
