Love the Question, Not the Answer

One of the things I’ve become increasingly interested in isn’t data itself, but how people respond when somebody questions it. Most organisations describe themselves as evidence driven, and many have invested heavily in platforms, dashboards and reporting processes intended to support better decisions. Yet I’ve noticed how quickly a perfectly reasonable question about a figure can change the atmosphere in a room.

Someone asks where a number came from, why a particular exclusion was applied or whether the same result would appear using a different reporting period. Instead of treating the question as an opportunity to understand the analysis more fully, the conversation can become defensive. The figure has already been presented, perhaps to senior leaders, and challenging its foundations starts to feel like challenging the credibility of the people who produced it.

I find that reaction interesting because it changes what data is being used for. Rather than helping us investigate reality, it becomes something we’re expected to defend. We stop exploring what might be true and start protecting what we’ve already said.

When confidence gets in the way

There’s nothing wrong with wanting confidence in our analysis. Organisations need information they can rely on, and endless debate about every measure would make it difficult to act. But confidence should come from understanding the assumptions behind a conclusion, not from discouraging people from questioning it.

Take a performance measure that has been reported consistently for several years. It may have an agreed calculation, a dependable pipeline and a familiar place in management reporting. Over time, people become comfortable with it. Then somebody new asks whether it still reflects the activity the organisation is trying to understand. The calculation hasn’t changed, but perhaps the service has, or the way people interact with it has evolved. Suddenly the question is not whether the dashboard is accurate, but whether the measure remains meaningful.

That distinction can be uncomfortable. A technically correct number may still encourage the wrong interpretation, and a definition that was appropriate when it was introduced may become less useful as circumstances change. A willingness to revisit those assumptions is not a sign of weak governance. I would argue it’s evidence that governance is working.

Curiosity is a professional discipline

The strongest analytical teams I’ve worked with haven’t been the ones most determined to produce an answer quickly. They’ve been the ones willing to ask what a question actually means, explore competing explanations and acknowledge what the available evidence cannot establish.

That requires technical skill, but it also requires a particular attitude towards uncertainty. An analyst needs to be able to explain how a figure was produced and still remain open to the possibility that it isn’t the right figure for the decision being made. A leader needs to be able to challenge an interpretation without treating the challenge as a criticism of the people involved. Both are easier to describe than to practise.

I keep returning to a phrase that captures this for me: love the question, not the answer. I don’t mean that answers are unimportant or that every decision should remain permanently unresolved. The point is that an answer should be something we can improve when better evidence or a better question becomes available, rather than a position to which we’ve become attached.

There’s also a responsibility on those asking questions. Curiosity shouldn’t become a way of avoiding decisions or undermining work without offering anything constructive. Good questions clarify assumptions, reveal uncertainty and help people understand what would change a conclusion. They create progress rather than simply prolonging disagreement.

What happens when we stop asking

The risk becomes more significant as organisations use AI to interpret data and recommend action. These systems can produce fluent, persuasive explanations that are easy to accept, particularly when they appear to be grounded in familiar internal information. The better they become at providing plausible answers, the more important it is that somebody remains willing to ask how those answers were reached and what might have been missed.

I don’t think a healthy data culture is one in which nobody disagrees about the numbers. It’s one in which disagreements can be examined without people feeling that their professional standing is under threat. Definitions can be challenged, methods can be improved and decisions can be revised when circumstances justify it.

Organisations will always need to act on incomplete information. Curiosity doesn’t remove that uncertainty, but it helps us remain aware of it. And in my experience, the people who make the greatest contribution are often those who can ask a difficult question without turning it into a contest over who was right in the first place.