The Analyst's Path

Handling questions in a presentation

Q&A is where the assessment happens. Your prepared part shows you can build something, and the questions show whether you understand it. It's also the part almost nobody rehearses, so a small amount of structure pays very well here. Every question belongs to one of five classes. Most of the skill is deciding which one, in the second of silence you take before answering.

7 min readPart of Delivering with Confidence: Rehearsal, Q&A, Remote and HybridOpen in course mode

Advertisement

Q&A is where the assessment happens. Your prepared part shows you can build something, and the questions show whether you understand it. It's also the part almost nobody rehearses, so a small amount of structure pays very well here.

Every question belongs to one of five classes. Most of the skill is deciding which one, in the second of silence you take before answering.

What are the five classes of question?

Answer is the default, and you should use it whenever it's available. The questioner has asked something you know, something on-topic that the room needs. Answer it directly, in one to two sentences, with the number, and then stop. Length is the commonest error here. A complete answer to a factual question rarely needs 90 seconds, and once an answer runs past two minutes it's turned into a second presentation nobody asked for.

Bridge is for a question adjacent to your argument, where a direct answer would be true but would strand the room in a side alley. Answer the adjacent thing straight, in one clause, then return to the claim. "Volumes are flat, yes, and the reason margin still fell 380 basis points is the input cost line, which is 210 of those points." You do answer, so a bridge isn't evasion. You just don't let the answer become the new topic.

Advertisement

Park is for a real question that's too big for the window. Name it as a real question, say where it'll be handled, and get agreement. "Whether we should be in quick commerce at all is the right question and it is a separate discussion; can I take it to the next board meeting with numbers?" A park needs the other party's assent, or it reads as a brush-off. You get that assent with a named venue and a date.

Defer is for a question of detail that only one person in the room cares about. "I have that by city tier; let me send it this afternoon rather than take the room through it." It's the cheapest class in the window and the one most often mishandled, because you'll be strongly tempted to prove you have the detail.

Commit is for a question you can't answer, and it gets its own section a little further on.

How do you plan the Q&A window?

Your mix of classes shapes the window. In the board Q&A worked example, the plan spends 205 of 375 planned seconds on the two full answers, or 54.7 percent. The remaining 45.3 percent goes across a bridge, a park, a defer and a commit. That's the right proportion. Give four of six questions full answers and the window will overrun.

Two mechanical habits carry a surprising amount of this. The first is to repeat or rephrase the question before you answer. It buys time and confirms you heard it, and it lets the rest of the room hear a soft-spoken question from the far end of the table. The second is to answer the person, then look away to the room. That stops a single questioner turning the Q&A into a private conversation.

Memorise the two-minute ceiling. An eleven-minute window holds five answers of two minutes, or eleven of one minute. Six questions in eleven minutes gives a mean of 110 seconds each. If the mean handling time you're planning exceeds two minutes, what you've prepared are appendices.

The hostile question: separate the challenge from the challenger

A hostile question has two components travelling together. One is a challenge, which is a claim about your work. The other is a challenger, who has a reason for making it in that tone in front of those people. Answer the challenge. Don't answer the challenger.

Take "Who signed off a 4.2 crore rupee overspend?" It's a challenge with an answerable core: the authority under which the money was committed. The tone carries something else, usually one of three things. Maybe the questioner wasn't consulted and is saying so publicly. Maybe they're protecting a position your analysis threatens. Or someone above them has asked them the same question, and they've nowhere to put it. Your tone of voice solves none of those three, and all three get worse if you match the temperature.

The mechanics are simple and hard.

Slow down. Under attack your instinct is to talk faster, and that reads as defensiveness even when the content is right. A two-second pause before a hostile answer is the highest-return habit you have under fire.

Restate the challenge in neutral words, and let the questioner correct you. "The question is whether the overspend was authorised, and by whom." If you've got it wrong, they'll tell you. Now you've turned an attack into a clarification. If you've got it right, the temperature has already dropped, because the sentence in the room is now yours.

Answer the substance in order: the fact first, the authority second, the consequence third. Then stop. A defence nobody asked for reopens the question you just closed.

Concede what's fair. A challenge that's right is a gift, because agreeing with it costs you nothing and buys you the rest of the room. "You are right that the base case assumes the competitor holds price, and we have no evidence for that; here is what the answer looks like if they cut 5 percent." That sentence has never damaged anybody's credibility.

Refuse only the part that's actually wrong, and refuse it on the evidence. Your authority won't carry it. "That is not what the data shows, and here is the line" beats "I have looked at this closely" every time. The second sentence asks the room to take your word for it at the very moment somebody has publicly declined to.

The one thing never to do is win.

A public victory over a senior person in front of their peers is the most expensive thing you can buy in a meeting room, and the bill comes to you.

The question you cannot answer

You'll be asked something you don't know. The room already assumes it, so the only variable is what happens next.

There's one structure that works, and it has three parts. Say plainly that you don't have it. Say what you do have that's adjacent, if anything, in one sentence. Then commit: what you'll send, to whom, by when, in that order and with an actual date.

"I do not have month-12 repeat rate. I have month-6, which is 41 percent, and the month-12 cohort only completes in November. I will send the month-12 number the week it closes, to the full board, with the cohort table behind it." Notice there's no apology in it, and no guess.

Three failure modes replace that structure, and every one of them is worse than the admission.

Guessing is the worst. A guessed number becomes a quoted number the moment somebody writes it down, and you'll spend a fortnight retracting it. If you must estimate a number live, say the word "roughly" and give a range. Then say what would tighten it.

The second is filling the silence with adjacent knowledge. It sounds like an answer for about twenty seconds. After that, everybody in the room understands that you haven't answered, and now they doubt the parts you did answer.

The third is the unbounded promise. "I will look into it" has no date, no recipient and no deliverable. It won't get done, and the questioner knows that as they hear it.

A commitment made in a meeting is a commitment logged in the minutes.

So write it down while you're standing there, out loud. The act of writing is itself the reassurance, and a commitment you didn't record is one you'll miss.

Advertisement

Part of the course

This page is part of Delivering with Confidence: Rehearsal, Q&A, Remote and Hybrid (PR6.01), in Phase 6: Investing philosophy, mental models and behavioural edge. The module's flashcards and its quiz are in the app.