How to answer tough prospect questions on a live call

A short pause is fine. It reads as consideration, even as competence — you are taking the question seriously. Somewhere around five seconds that flips. The silence stops being thought and starts being information, and the people on the call begin supplying their own explanation for it.

The practical consequence is that you are not deciding whether to answer well. You are deciding, inside five seconds, which of three moves to make. Knowing the three in advance is most of the skill.

Obvious, but worth stating precisely: answer it if you know the answer, not if you think you probably know it. The failure mode here is the half-remembered specific — a retention period, a rate limit, a compliance certification you believe you hold.

Those are exactly the answers that get written down. A prospect asking about data residency is not making conversation; they are filling in a form that legal will read. Being approximately right there is worse than being honestly unsure, because the correction arrives later, in writing, and looks like something else.

This is the most underused move in sales, and the one that costs least when used well. The sentence is short:

“I don't want to guess at that one — it's important enough to get exactly right. I'll have it to you by end of day tomorrow.”

Three things make it work. You decline to guess, which signals that your other answers were not guesses either. You name why — the question matters — so it reads as rigour rather than ignorance. And you give a specific time, which converts an open loop into a commitment.

What ruins it is vagueness. “Let me check and get back to you” with no deadline is the phrase prospects have heard a hundred times, and it lands as a polite nothing.

Sometimes the question rests on a premise that does not apply — someone asks how you handle a limitation that your architecture does not have. Correcting the premise is legitimate and often the most useful answer available.

The abuse of this move is answering an adjacent question you prefer. Everyone notices. Someone who has asked a specific technical question and received a confident answer to a different one now believes you are managing them, and the rest of the call is spent recovering from that rather than selling.

Filling the pause with words while you think. Prospects hear it, and it reads as improvisation rather than knowledge. Deflecting to a colleague who is not on the call. And promising to follow up on something you have already promised to follow up on, which tells the room that the first promise meant nothing.

The move that compounds is boring: keep a list. Every question that caused a pause, written down the moment the call ends, with no judgement about whether you should have known it.

Once a week, find the written answer for each one. Two things usually emerge. Most teams find the same six or so questions account for most of the damage, which makes the problem far smaller than it felt. And some questions have no documented answer anywhere — the gap was never yours, it was the company's, and the call simply surfaced it.

A real-time assistant can close the five-second window by putting the relevant passage on screen while the question is still being asked. That is genuinely useful — but only if it answers from your own documents.

A general model asked about your encryption or your uptime commitment will produce something confident and invented. That does not solve the problem in this article; it industrialises it. The useful version is constrained to material your company has already approved, so what appears on screen is a sentence you could have read aloud from the whitepaper if it had been open.

It still will not tell you when to decline, when to escalate, or when the honest answer is that you do not support something. Move two stays part of the job.