By the Time It's Obvious, the Window Has Usually Moved.

Igor Ansoff — the strategist behind the Ansoff Matrix — published "Managing Strategic Surprise by Response to Weak Signals" in California Management Review in 1975. His argument was that most damaging strategic surprises aren't actually sudden; the information was available earlier, just in a form too vague and unfamiliar to act on with confidence. Ansoff proposed matching an organization's response to the strength of the available signal — treating a vague, early indication differently from a fully confirmed threat, rather than waiting for every signal to reach the same threshold of certainty before responding at all.

The theory has real critics — some researchers have noted the empirical evidence for how reliably weak signals can actually be identified in advance is thinner than the theory's popularity suggests. That criticism is worth taking seriously, and it points at exactly the right problem: the theory says organizations should read weak signals, but it's much less specific about how, operationally, to do that without drowning in noise or waiting too long for confirmation. Ansoff's real insight was that the signal usually existed earlier than the response did — his framework just doesn't fully solve how an organization tells a real early signal apart from noise before the fact, only after.

Later research on weak signals, including work by Ilmola and Kuusi, found organizations frequently fail to notice or act on a weak signal specifically when it conflicts with existing assumptions or established decision routines — not because the signal wasn't detectable, but because acting on it required contradicting something the organization already believed. That's a harder problem than data availability. It's an organizational one.

The instinctive response to missing a signal is to collect more of them. Richards Heuer's Psychology of Intelligence Analysis — written for the CIA, still standard reading in intelligence and analytical training decades later — argues directly against that instinct: Heuer found that the standard prescription of "more and better information" rarely fixes weak analysis, because the failure usually isn't a shortage of data. It's how the data already on hand gets processed.

Heuer's research catalogs specific, well-documented cognitive failure modes that get worse, not better, as volume increases: confirmation bias, where ambiguous evidence gets read as supporting whatever the analyst already believed; anchoring, where an early impression distorts every piece of information that follows it; and the availability heuristic, where whichever explanation comes to mind most easily gets treated as the most likely one, regardless of actual likelihood. Heuer's central finding is blunt — simply knowing these biases exist does very little to prevent them. What helps is structured technique, applied consistently, by people experienced enough to apply it under pressure rather than default to the easy read.