The atom of synthesis is the difference between a finding and an insight, and the two are different species. Making a finding bigger or better never turns it into an insight. A finding is a true statement about the data: "Revenue grew 14% this year." "Segment B has a 4% margin." "Churn in the mid-market cohort is 22%." Each is correct and each was hard-won. Delivered on its own, each one earns the same reply: so what, is that good, why did it happen, what do you want me to do about it? An insight answers that reply. It states what the finding means for the decision. "Revenue grew 14%, and critically it was all volume at flat price, so the company is winning units without discounting, which confirms real pricing power rather than a promotional sugar-high." Same fact, transformed. The finding reports the number, and the insight tells the client what to believe and do because of it.
Advertisement
What counts as an insight?
Once you have the test that separates them, you can't unsee it. An insight is actionable or belief-changing. A finding is merely true. A finding just sits there. It's correct and useless. An insight moves something: it changes what the client should do, or what a smart, informed client already believed. If a statement can be true and leave the decision where it was, it's a finding, however much analysis produced it. So "revenue grew 14%" is a finding, and "the 14% was all volume, so the moat is real and we should defend price, not chase share" is an insight. The first changes nothing, while the second changes the pricing strategy. Beginners fill decks with the first kind believing they're delivering the second, because both took the same effort. But effort isn't insight, and the client can feel the difference in the first four minutes. A page of findings is homework handed in. A page of insights is analysis that has turned into advice.
How do you test for an insight?
Keep three practical tests for whether you've really reached insight, because "so what?" can be answered lazily as easily as it can be answered well:
- The action test. Does the statement imply, or point straight at, something to do or decide, or to believe differently? If nothing follows from it, you're still on a finding.
- The surprise/movement test. Would a smart client who already knows their own business be moved by this? Or did they know it already? "Our biggest market is our biggest market" passes no test. The insight is in what the analysis revealed that the client's own intuition missed: the exception, the de-averaged truth, the second-order effect.
- The decision test. Does it point at the question you were hired to answer? An insight about something the client didn't ask about, however clever, is a distraction dressed up as insight. The decision is what keeps synthesis on track.
An observation that fails all three is what practitioners call "interesting but inert": true, sometimes even surprising, and utterly beside the point. Killing these observations is as much a part of synthesis as lifting your findings up. It's also harder, because they're your favorites.