What makes a management foundations question actually return information?

A manager's main instrument is the question, and most questions are built so they cannot return bad news. "Everything on track?" has one comfortable answer and one costly one, and you have already indicated which you would prefer.

This page is about the construction of the question rather than a list to memorize. Change the wording and you change what you find out, which is the cheapest improvement available to anyone running a team.

What to take away

What makes a question return real information

Four properties do most of the work, and each one is easy to check before you open your mouth.

It asks about a specific event, not a general condition. "How is the project going?" is answered from mood. "What was the last thing that slipped, and what caused it?" is answered from memory, and memory contains detail.

It does not carry your preferred answer. "You're happy with the new process, right?" has been answered by the person asking. Strip the tag and the adjective: "What is the new process making harder?" A question shaped around the answer you expect collects agreement rather than information, which is confirmation bias with a microphone attached.

It is answerable without a confession. "Why hasn't this been done?" makes the person the subject. "What is this waiting on?" makes the work the subject. Same information, different cost to supply it.

It has a scope small enough to answer honestly. "Do you have capacity?" invites an estimate nobody can make accurately. "If I gave you this on Thursday, what would move?" gets a real answer, because it names a trade rather than a feeling.

Questions that fail in predictable ways

The pattern in the right column is the same each time. Ask for something the person can look up in their own recent experience, and the answer stops being a performance.

The second question

Most useful information arrives after the first answer has been given and not accepted as final. Not challenged: not accepted as final.

There are three moves worth having ready. "What else?", repeated once, because the first item is the socially cheapest one and the third is often the real one.

"What would have made that easier?" converts a complaint into a change you can act on. "What am I missing?" is the only one here that reliably gets you criticism of your own decisions, and only if you have acted on the answer before.

Then stop talking. Counting to five in your head after a question feels absurd and produces more than any rephrasing. People fill silence, and managers fill it fastest, usually with the answer they were hoping for.

Asking about your own performance

"Any feedback for me?" gets nothing, every time, and it should. The person is being asked to volunteer criticism of someone who influences their pay and their work, on the spot, with no preparation.

What works better is narrow, specific and about a thing rather than about you. Ask what you should have decided sooner. Ask which meeting of yours they would drop. Ask what you asked for last quarter that turned out not to be worth doing. Each names a candidate answer, which makes the reply an edit rather than an accusation.

Then act visibly on one item, quickly, and say that is why you are doing it. Until the first person who spoke up sees a result, you are collecting politeness. This is the mechanism behind most of what makes team management work at all, and no amount of asking substitutes for it.

Questions for particular moments

A new joiner, in the first month. "What has been confusing that nobody warned you about?" This window closes. After about eight weeks they will have normalized every oddity in your team, and the information is gone permanently.

After something went wrong. "What did it look like from where you were standing, before you knew how it turned out?" Knowing the ending makes the path to it look inevitable, an effect reliable enough to be called hindsight bias, and it rewrites the story into a sequence of obvious errors. The pre-outcome version is the one that tells you what to change.

When someone brings you a decision. "What would you do if I were away?" This distinguishes a genuine escalation from a habit of checking in, and the answer tells you where the delegation boundary actually sits.

When you inherit a team. "What is the thing everyone knows that a new manager usually takes six months to find out?" Ask it individually, not in a group. The management foundations guide covers the rest of that first period.

When work is going well. Ask anyway, and ask what is fragile about it. Things that are working are the least examined and the most likely to be quietly held together by one person.

What to do with the answers

Two rules keep the whole practice from decaying.

Write down what you were told and what you committed to, during the conversation rather than after it. The fastest way to make people stop answering honestly is to agree to remove an obstacle and then not do it. Two rounds of that and the questions become ceremony.

Separate finding out from evaluating. A question asked to learn something and a question asked to build a case are different activities, and people can tell which one they are in.

When you move toward anything with formal consequences, make that shift explicit; the procedure belongs to your HR function, not to a conversational technique. The distinction is in performance management.

For the underlying obligations these questions serve, see management foundations; for what happens when the answers reveal that a change is needed, change management is the next step.

Common questions

Is this not just manipulation with better wording?

The test is whether the question survives disclosure. Telling someone "I ask what would move because a yes to extra work is usually not a real yes" makes the question work better, not worse. A technique that stops working once explained is a different thing, and you should not use it.

What if someone still tells me nothing?

Then either the honest answer is expensive for them, or nothing is wrong. Assume the first for a while. Ask about someone else's obstacles rather than their own, ask in writing so they can answer without you watching, and check what happened the last few times a person on your team raised something uncomfortable. The precedent is usually the explanation.

How many questions is too many in one conversation?

Three that get real answers beat twelve that get short ones. If you are moving through a list, you are conducting an audit, and people answer audits defensively no matter how the questions are worded.