Saying I Don't Know on Sales Calls: Drills

What should a rep say when the buyer asks something they can't answer?

The rep says the gap out loud in one sentence, then restates the question, names a dated follow-up, and returns to discovery, without guessing at any point. Answering questions you can't answer in sales is a five-check behavior: four things the rep does and one thing the rep never does. Each check is scoreable from the transcript.

The sequence: (1) state the gap plainly, I don't know that. (2) Restate the question so the buyer hears it was understood, You're asking whether the export keeps field-level history, not just the final record. (3) Commit to a named follow-up with a date, I'll get that from our product lead and email you the answer by Thursday noon. (4) Return to the discovery thread, Back to month-end reconciliation: who touches that today?

The fifth check is a constraint that runs through all four moves: no guess. No I think it's probably and no it should work. You score it by the absence of any hedged estimate or inferred capability in the response.

The restatement matters more than reps expect. It proves the rep listened, and it gives whoever researches the answer a precise question instead of a vague one. The return to discovery matters because a rep who lets the call stall on the gap hands the buyer control of the agenda.

Each move is short. The whole sequence runs a few sentences. If the rep needs a paragraph, the rep is apologizing or stalling.

A plain admission paired with a dated commitment tells the buyer that the rep's other answers can be trusted.

Why does a guess cost more than the gap?

A guess costs more because the buyer can check it and the rep usually cannot take it back. A gap costs one follow-up email. A wrong guess about a product detail, a reference customer, or a roadmap item can surface in a security review, a procurement call, or a signed contract.

We hold the position that live pipeline is the most expensive place to learn this. A rep who bluffs for the first time on a real buyer is practicing on the deal. Rehearse the admission in scored practice first, so the sentence I don't know that is already automatic when the buyer's question lands.

The opposing view deserves a fair hearing. Some managers say strong reps read the room and improvise a confident answer, and that admitting ignorance signals weakness. That works when the rep actually knows the answer. It fails when the confidence is covering a hole, and a buyer who later finds the hole may discount everything else the rep said. The drill does not forbid improvisation. It makes the admission available so improvisation is reserved for what the rep understands.

This is also why the drill is a sales follow-up commitment drill as much as an honesty drill. The admission is only half the behavior. The dated promise is the other half, and it is the half managers can check against what happened next.

A wrong guess converts one missing answer into doubt about every answer the rep has already given.

The rubric row and pass bar

Score the behavior as one row with five checks. Every check must be verifiable from the transcript without listening to tone. If you use XL Roleplay, flags in the report link to the exact transcript moment; on any other setup, mark the timestamp yourself.

The pass bar we recommend: all five checks present in a single response, on every scenario in the session. One missing check fails that attempt. Treat this as our recommendation for your rubric, not a standard from research.

The no-guess check is a constraint, so you score it by absence: the transcript contains no estimate, hedge-claim, or inferred capability anywhere in the response.

A rubric row passes only when a manager can verify every element from the transcript without hearing the audio.

What does a failing bluff or stall sound like?

A failing bluff sounds fluent and specific about something the rep does not know. A failing stall sounds helpful and specific about nothing. Both fail the row, and you need to hear both so the debrief can name them.

The bluff: buyer asks, Does your audit log capture who changed a pricing field? Rep answers, Yes, absolutely, every change is tracked. No gap stated, no source, and the rep has no idea. Scored: fails the state-the-gap and no-guess checks.

The hedge-guess: I believe it does, I'd have to confirm. The rep sounds honest, but the claim is already on the record. Scored: fails the no-guess check. This one is easy to miss because it contains the word confirm.

The stall: Good question, let me circle back on that. No gap named, no restated question, no owner, no date. Scored: fails state-the-gap, restate-the-question, and named-follow-up. A stall sounds helpful but names no owner, no deliverable, and no date, so the buyer has nothing to hold the rep to.

The abandoned thread: the rep delivers a clean admission and a dated promise, then goes quiet or jumps to the next slide. Scored: fails return-to-discovery. The buyer was mid-story about a problem and the rep left it.

How do you run the drill this week?

Run it as a 30-minute block: 5 minutes of setup, 15 minutes of drill, 10 minutes of debrief. We recommend you spend about 10 more minutes beforehand picking the scenario from a live deal and choosing the rubric row. Put it on the calendar as protected coaching time, separate from the pipeline review.

Use three scenarios in one session, one per question type. Product detail: the buyer asks about a configuration your rep has never touched. Customer reference: the buyer asks for a named customer in their industry the rep cannot name. Roadmap: the buyer asks whether a missing capability ships before their renewal date. For reference-specific practice, see Customer Reference Call Drills for 2026. For missing-capability pressure on demos, see Feature Gap Objection Drills for Demos.

Play the buyer, or assign an AI persona, and tell the buyer to ask the question mid-discovery, after the rep has just heard a pain point. That placement is the test, because the return to the discovery thread has to land on something real. Tell the buyer to push once: Just give me your best guess. A rep who holds the line on the first push passes the no-guess check under pressure.

On roadmap questions, the follow-up has a different shape. The rep cannot promise a feature, so the commitment is, I'll find out what I'm allowed to share and tell you by Friday what I can and can't say. Score that as a pass.

Run length is a recommendation. Adjust rep counts to your team, but keep the three scenario types and keep the push.

What is the re-run rule?

A failed attempt is re-run, in the same session when time allows, on the same question type, before the debrief widens to anything else. If the session runs out, schedule the re-run for a named day this week. The rule is the point of the drill: a rep who fails and hears only a lecture has not practiced the behavior, and a rep who repeats it under the same pressure has.

Debrief in this order. First, play or read back the one failing moment and ask the rep to diagnose it: Which of the five checks did you miss? Let the rep answer before you say anything. Then name one behavior, not three, even if the transcript shows several misses. Pick the earliest check that failed. Then re-run. Our guide to one behavior, re-run debriefs covers the sequence in full.

If the rep fails the re-run, schedule a second one for a specific day this week rather than looping in the room. A rep who has failed twice in a row may be missing a script line. Hand over the four lines and have the rep say them aloud once: I don't know that. Then You're asking whether... Then I'll get that from [owner] and send it by [day]. Then Back to what you said about... For teams that gate on this row, the retake policy guide covers how many attempts to allow.

A drill with a pass bar, a failing sound, and a scheduled re-run keeps the standard the same for every rep.

Where does this drill fall short?

The drill certifies the sentence, not the knowledge behind it. A rep can pass every attempt and still be unprepared on the product. If a rep reaches for the admission on questions the rep should know cold, that is a preparation gap, and the fix is product study, not more drilling. Honest admission of a gap is a skill for the rare question; a rep who needs it constantly has a preparation problem.

The row also cannot verify that the follow-up happened. The transcript shows the promise. Only your CRM, the email thread, or the buyer's next call shows delivery. We recommend checking delivery on live deals for the first few weeks after you introduce the drill, and treating a missed date as a coaching moment for that rep, scheduled in its own 15-minute 1:1 and not folded into the pipeline review.

Finally, some buyers read any admission as weakness, and the drill will not change that. The trade-off is deliberate. In our view, a rep who admits one gap and follows through is better placed than a rep who is never wrong until the contract review. If your buyers are the exception, say so in your rubric notes and adapt the scenario persona, not the standard.

Corrections or disagreements with this guide can go to coach@roleplay.xl.net.