Ask Questions That Come With Options

There is a habit I developed somewhere between running classrooms, running investigations, and running systems work, and I now consider it one of the most useful moves in a leader's kit. When I ask a diagnostic question, I ship the candidate answers with it.

Not: why do you think submissions dropped? Instead: submissions dropped, and I can think of three explanations, people do not trust where the input goes, people do not know the channel exists, or something in the structure itself changed. Which is closest, or is it something I have not listed?

The difference looks cosmetic. It is not. Those two questions make completely different requests of the person answering, and they return completely different data.

The blank page is expensive

An open-ended question hands someone a blank page, and blank pages carry costs that leaders, who spend all day being asked things, tend to forget. The answerer has to generate hypotheses from scratch, on the spot, in front of you. They have to guess what altitude you want, a sentence or an analysis? They have to gauge the politics of every candidate answer before saying it, because they do not know which possibilities you have already considered, or which one you are hoping for, or which one indicts someone in the room.

So what comes back is often the safest available noise: some version of it is probably a lot of things. That answer is not evasion, mostly. It is a rational response to unpriced risk. The person had real information, and the blank page made it too expensive to retrieve and too dangerous to volunteer.

A question with options collapses all of those costs at once. Generation is done, three hypotheses are on the table. Altitude is set, pick one or improve on it. And the politics are defused, because every option on the list is one you were visibly willing to say out loud yourself. You have pre-cleared the answers. The person can now do the one thing only they can do: match the options against what they actually know.

Options show your work, and showing your work invites correction

The deeper function is dialogical. When I offer my three candidate explanations, I am not just making answering cheaper. I am exposing my current model of the situation, in checkable form. The person now knows exactly what I think might be going on, which means they can tell me exactly where I am wrong.

That is the move Chris Argyris spent a career trying to get leaders to make: pairing advocacy with inquiry, stating your view and your reasoning, then genuinely inviting others to test it (Argyris, 1991). Most leaders do one or the other. They advocate, here is what is happening, or they inquire from behind glass, what do you all think, revealing nothing. The question-with-options does both in one breath. Here is my thinking, such as it is. Attack it.

And notice the escape hatch, which is not optional: or is it something I have not listed? The options are scaffolding, not walls. In my experience the fourth answer, the one nobody upstairs had imagined, arrives more often when three imagined ones are already on the table, because the list proves that specific, disconfirming answers are what is wanted. The constraint, paradoxically, is what licenses the surprise.

Where the tool belongs, and where it does not

I hold one boundary on this. Constrained questions are diagnostic and operational instruments. They shine when something concrete needs locating: why a number moved, which of three dates works, where a process is snagging. They also shine downward across a power gap, precisely because the power gap is what makes blank pages expensive, and the options are how the powerful party pays the toll instead.

But dialogue, in the full sense, exploring how a situation looks from other seats before any decision forms, needs open questions and the patience they cost. If I only ever asked questions with options, people would only ever explore terrain I had already mapped. So the practice is matched to the meeting. In the understanding phase, blank pages, protected time, no menu. In the locating phase, question plus candidates plus escape hatch, and act on what comes back.

For example, this habit compresses routine operations to almost nothing. A scheduling conflict that could generate four emails of open-ended confusion resolves in one: this date reads wrong to me, was it correct, did the event move, or is this a mistake we need to adjust for? The person closest to the facts picks door two, adds a sentence, done. Multiply that by every diagnostic exchange in a week and you get something better than saved time. You get an organization where questions feel answerable, and answerable questions are the ones that actually get answered.