What to say when you do not know the answer on a live call
Never guess and never stop at "I do not know". Use three parts: say plainly that you do not have it, say what you do know that is adjacent, and commit to a specific time you will come back with the answer. Said in one breath, it reads as competence rather than a gap.
- What guessing actually costs
- Why a bare "I do not know" also loses
- The three-part response
- It changes by call type
- Phrases worth having ready
- When it keeps happening on the same call
- The part that actually builds the credibility
- The real difficulty is the two-second window
What guessing actually costs
The instinct is to produce something, because silence feels like failure and "I do not know" feels like losing. So people approximate, hedge, or talk around it.
The arithmetic is worse than it looks. A wrong answer on a sales call surfaces later, usually in front of more people, and what gets remembered is not the mistake but that you were confident about it. On a support call it can mean a customer acts on bad information. In an interview a fabricated answer collapses at the first follow-up.
The specific damage is not the error. It is that every correct thing you said afterwards now carries a discount, because the listener has learned they cannot tell your confident from your uncertain.
Credibility is not built by knowing everything. It is built by the listener being able to trust the difference between when you know and when you do not.
Why a bare "I do not know" also loses
The opposite failure is real too. "I am not sure" delivered flat, with nothing after it, hands the conversation back to the other person with nothing in it.
It reads as a dead end rather than honesty, because you have reported a gap and taken no responsibility for closing it. Worse, it invites the follow-up nobody wants: "well, who would know?"
The problem is not the admission. It is stopping there.
The three-part response
Say all three in one breath, without a pause between them. The pause is what makes it sound like a confession.
- Name it plainly. "I do not have that number in front of me." Short, specific, no apologising, no hedging language.
- Give the adjacent thing you do know. This is the part people skip and it does most of the work. You almost always know something useful near the question.
- Commit to a specific return. Not "I will find out" but "I will have it to you by Thursday morning." A time and a channel.
What it sounds like
"I do not have the exact migration timeline for a team your size in front of me. What I can tell you is the two things that drive it are how many custom fields you have and whether you need historical data moved. I will get you a real number by Thursday morning."
Nothing was invented, the prospect learned something genuinely useful, and there is a dated next step. That is a better outcome than a guessed number, including in the case where the guess would have been right.
It changes by call type
| On a... | The risk | Emphasise |
|---|---|---|
| Sales call | Guessing a number you cannot honour | The adjacent knowledge, then the dated return |
| Support call | Customer acts on bad information | Naming it plainly and fast; accuracy beats speed |
| Job interview | Fabrication collapsing at the follow-up | How you would find out, and your reasoning |
| Negotiation | Conceding ground by filling silence | Naming it, then returning the question |
The interview case is different
In an interview, "I do not know" is frequently the question. Technical interviewers ask things you cannot know to see what you do next. The right answer is your reasoning: "I have not worked with that directly. My instinct would be to start by checking X, because in the analogous case with Y that was usually the cause." You are being assessed on method, not recall.
The negotiation case is also different
Do not rush to fill it. "I would need to check that" followed by silence is a complete move. The pressure to fill a gap is exactly the pressure that produces concessions nobody asked for.
Phrases worth having ready
Because the moment arrives fast and you will not compose anything good from scratch.
- "I do not want to guess at that, so let me get you the real number." Explicitly refuses the guess, which reads as discipline.
- "That is outside what I work on directly. [Name] owns it and I can have them answer today."
- "I know the shape of the answer but not the detail, and the detail is the part that matters to you."
- "Good question, and I am not certain. Here is how I would find out." For interviews.
- "Let me check rather than tell you something wrong." Ends the exchange cleanly.
And what to stop saying
- "I think it is probably around..." Hedged guessing is still guessing, and the hedge disappears from the listener's memory.
- "I will find out." No date, so no commitment.
- "That is a great question." Universally recognised as stalling.
- Silence longer than about three seconds. It reads as being caught out, whatever you say next.
When it keeps happening on the same call
Once is fine. Three times in twenty minutes is a different problem, and pretending otherwise makes it worse, because by the third one the other person has definitely noticed even if you have not.
Name the pattern before they do. "You are asking the right questions and I am hitting the edge of what I can answer usefully. Let me bring in someone from the implementation side rather than keep half-answering." That is a stronger move than the third improvised response, and it reframes the gap as a scoping issue rather than a competence one.
Afterwards, treat the repeats as data. Three unknowns clustered in one area is telling you where your brief is thin, and it is almost always the same area across calls: pricing edge cases, integration specifics, security questions. That is a fixable gap, and fixing it once removes the problem from every future call rather than just this one.
The part that actually builds the credibility
None of this works if you do not come back. The move is not the admission, it is the admission plus delivery, and the second half is where people lose the goodwill they earned.
Write the promise down during the call, not after. Send it early rather than on the deadline. And send it even when the answer is unhelpful, because "I checked and we cannot do that" delivered on Thursday as promised is worth more than a vague yes.
Done twice with the same person, this becomes the thing they trust you for. Buyers and customers deal with people who overclaim constantly; someone who reliably distinguishes what they know from what they will check is unusual enough to be memorable.
The real difficulty is the two-second window
Everything above is easy to agree with and hard to execute, because the gap between the question landing and your answer starting is about two seconds wide. In that window you do not reason, you retrieve, and the default retrieval is the hedge.
Two things move it. Rehearsal, until the three-part response is the automatic one rather than the considered one. And having the adjacent facts visible, so the middle step is available rather than something you have to dig for. That is the job Unblank does on live calls: your own material on screen the moment the other person stops talking.
If your failure mode is freezing rather than guessing, what to say when you blank out on a sales call is the closer read. For support specifically, see how it works on support calls.
Frequently asked questions
What should I say when I do not know the answer on a call?
Three parts in one breath: name plainly that you do not have it, give the adjacent thing you do know, and commit to a specific time you will come back with the answer. Pausing between them is what makes it sound like a confession.
Is it bad to say "I do not know" in a sales call?
No, but stopping there is. A bare admission hands the conversation back with nothing in it. Paired with adjacent knowledge and a dated follow-up, it reads as discipline rather than a gap.
What should I say in an interview when I do not know?
Give your reasoning. "I have not worked with that directly, my instinct would be to start by checking X because in the analogous case Y was usually the cause." Interviewers often ask unknowable questions specifically to see your method.
Is guessing ever better than admitting it?
No. A wrong answer surfaces later, usually in front of more people, and what gets remembered is that you were confident about it. Every correct thing you said afterwards then carries a discount.
What phrases should I avoid?
"I think it is probably around" is hedged guessing, and the hedge drops out of the listener's memory. "I will find out" has no date so it is not a commitment. "That is a great question" is universally recognised as stalling.
How do I get this right in the moment?
You have about two seconds before the reflex fires, so reasoning is not available. Rehearse until the three-part response is automatic, and keep the adjacent facts somewhere visible so the middle step does not depend on recall.
Keep the facts you actually have on screen, so the gap you admit to is the only one. See pricing or start with the free trial.
Download Unblank free