You've been in this moment. Something is wrong - with a hire, a deal, the direction the business is heading - and you know it, but you can't yet say what.
The Instinct-Ahead-of-Evidence Moment
Research on expert judgement supports this precisely. Gary Klein's work1 on recognition-primed decision making describes how experienced practitioners detect anomalies: they recognise a situation as a type they've encountered before, and that recognition fires before explicit reasoning follows.
This happens more to experienced leaders than to newer ones. Experience doesn't make you clairvoyant; it gives you more patterns.
The pressure this moment creates is the pressure to resolve it in one direction or the other. You either act on the instinct - and risk being wrong about something consequential, or you wait for evidence, which risks watching the thing you sensed play out while you were still waiting to be sure.
Most leaders who've had this experience and had it handled poorly remember two things. They remember the advisor who said "trust your gut; you know your business better than anyone," which felt supportive until the instinct turned out to be wrong. They also remember the advisor who said "come back when you have something concrete," and watching the thing they sensed becoming a reality.
Two Mentoring Failure Modes
Premature validation and reflexive dismissal share a common root: they both treat instinct as binary. Either the leader can be trusted - in which case validate, or they can't - in which case require evidence. The question "is this instinct reliable?" is treated as answerable and is answered, quickly, in one direction or the other.
But instinct isn't binary. It's information at an early stage of interpretation. The leader has noticed something. The mentor's job is to help them find out what.
The validating mentor, even with the best intentions, short-circuits that process. They endorse a conclusion before the observations underneath it have been surfaced. This is a problem for three reasons. The leader may be wrong; and if they are, they've made a consequential decision on an unexamined belief. Even if they're right, they won't know what they were actually responding to, which means they can't calibrate that capacity over time. And the mentor has put themselves on record in support of something untested.
The dismissing mentor treats the leader's pattern recognition as though it isn't evidence - as though the observations the leader has made don't count until they've been converted into a form the mentor can accept. This misunderstands what is happening. The leader does have something. They have noticed specific things. What they haven't done yet is articulate those observations and separate them from the interpretation.
What Good Mentoring Looks Like at This Moment
There are three moves that support this dilemma. They're sequential, but not a formula; they may run across one conversation or several.
Legitimise the signal without endorsing the conclusion. This is a precise distinction, and it matters enormously in practice. "That is worth taking seriously" is not the same as "you're right." The leader needs to hear the former: that their pattern recognition is real, that it counts, that the mentor is taking it seriously. They do not need to be handed a conclusion that hasn't yet been examined. The immediate risk is that the leader collapses in one of two directions - either self-dismissal ("it's probably nothing") or premature action ("I need to act on this now"). The mentor holds the space between those two.
Separate the observation from the interpretation. The leader comes in with a conclusion. The mentor needs to excavate what's underneath it.
Consider a founder with a read on a hire three months in - one of the first senior appointments, consequential and still settling in. Everything on paper looks fine. But something feels wrong. When asked what they have actually noticed - not the conclusion, but the observations - the answer might be: three late arrivals to Monday briefings without explanation; quieter than usual in the last two group sessions; no challenge to the Q3 pricing decision, when this person has pushed back on every significant call before now. Those are observations. They're not a conclusion. The interpretation may or may not prove correct. But separating what the leader has seen from what they have concluded is where the interpretation happens – and almost every leader, going through this process, finds they have noticed considerably more than they realised.
Set a test for the instinct. The question "should I trust my instinct?" cannot be answered. The question "what would I need to see to know whether my instinct is right?" can be. What evidence would confirm the read? What evidence would revise it? What would genuinely change your mind?
A test converts the instinct from something to agonise over into something to investigate. It also guards against motivated reasoning: a leader who can state in advance what would shift their view is less likely to unconsciously select for confirming evidence over the days that follow.
For the founder above: schedule direct conversations about expectations with the hire; ask open questions of the people she's worked most closely with; observe whether the behavioural pattern holds over three further weeks. The instinct is neither acted on nor dismissed.
After the Instinct: Building Judgment Over Time
Every instinct-ahead-of-evidence moment, handled well, is a small exercise in calibration. Every one handled badly is a small erosion of it.
Instinct calibration develops the way other skills do: through deliberate practice and feedback. Each debrief sharpens the leader's understanding of their own pattern recognition; what it tends to be responding to, where it can be trusted, where it has a track record of misfiring.
The debrief questions are specific. If the instinct was right: what did you actually notice that led you here? Which observations turned out to be the signal, and which were noise? What were you picking up on that you may have been inclined to dismiss? "I was right" is less useful than "I was right because I picked up on three specific behavioural cues I now know to take seriously".
If the instinct was wrong, the debrief is equally important. What were you pattern-matching to? Was there a previous situation that didn't actually apply here? Getting this wrong without examining it is how blind spots form. The recognition system works on the data it's been given. Unexamined false positives come back around.
A mentor who consistently debriefs instincts, both right and wrong, is building something. The leader's confidence in their own judgment becomes grounded, not in the feeling that they should trust their gut, but in an accurate understanding of when their pattern recognition can be relied upon and when it should be tested.
Think of the last time you had a strong read that turned out to matter. What did you do with it? And what did you learn about your own judgment from how it resolved?
The instinct-ahead-of-evidence moment doesn't need to be solved. It needs to be held. Good mentoring at this moment is a specific practice – and, over time, it's what turns a leader's pattern recognition from a source of uncertainty into a source of usable insight.
References
- Klein, G. (1998). Sources of Power: How People Make Decisions. MIT Press. https://mitpress.mit.edu/9780262611466/sources-of-power/
Ady Coles works as a thinking partner and mentor to leaders and teams navigating complexity. His work centres on judgement, perspective, and the often-invisible work of translation - helping people understand their role in the system, make better decisions, and operate with confidence in uncertain environments.