Stacked Questions in Sales Calls: A Drill

What is a stacked question in a sales call?

A stacked question is one rep turn that contains two or more questions asking for different information. The rep speaks once and the buyer hears a bundle. In the drill, the buyer persona is set to answer only the cheapest part.

Here is the verbatim shape. The rep says: "What's driving this project, who else is involved, and what does your timeline look like?" That is three answer slots in one turn. The buyer says: "We're hoping to start in Q3." The timeline got answered. The driver and the stakeholders did not, and the rep moves on because the buyer did say something.

Stacking is not limited to lists. Tag stacks are also failing lines: "How is the team handling renewals today? Is that mostly manual?" The second question narrows the first before the buyer has answered it, and it hands the buyer a yes to give instead of a description. Option stacks count as well: "Is it more of a cost issue or a speed issue or a compliance thing?"

A stacked question gives the buyer a choice of which question to answer, and the others may go unanswered.

Why do buyers skip the hard question in a stack?

Assume a buyer may skip the hard question in a stack, and check the transcript for whether each question received an answer. Do not read a skip as evasion. The turn the rep made ambiguous is the thing to fix.

Stacking is especially costly when a methodology depends on order. The Sandler pain funnel runs from the surface issue, to its concrete cost, to how the buyer feels about it. A rep who asks "What did that cost you, and how do you feel about it?" has collapsed two steps into one turn. The buyer may answer only the feeling, and the cost figure can drop out of the transcript. The ordered sequence was the point, and the stack removed it.

A skipped question leaves no trace in a recording unless someone matches each question the rep asked to an answer the buyer gave.

How does a manager count stacked questions from a transcript?

A manager counts stacked questions by marking each rep turn, defining a turn as everything the rep says between two buyer utterances, then flagging any turn with two or more questions that seek different information. No call memory is needed. The transcript is the only evidence.

Use these counting rules so two managers get the same number.

Rule 1. Mark only discovery-stage turns. Ignore the opener, the agenda, and the close.

Rule 2. Count answer slots, not question marks. "What did that cost you? How much, roughly?" is one slot restated, so it scores as a single question with a rewording note. "What did that cost you, and who else felt it?" is two slots and a stack.

Rule 3. A preamble before one question is fine. "That makes sense. What did that cost you?" passes.

Rule 4. Menu and tag questions count as stacks if they carry an extra slot or supply the buyer's answer.

On XL Roleplay scorecards, flags link to the exact transcript moment, so a manager can jump straight to a marked turn. A highlighter on an exported transcript works the same way for any tooling.

Record the result per rep and per scenario: stacked turns out of discovery-stage turns. Do not blend scenarios into one number.

Two managers counting answer slots in the same transcript should reach the same number; if they do not, the rubric row needs recalibration.

What does the rubric row look like?

The rubric row is one line, scored per turn and summarized per call. Name it "One question per turn." Its observable behavior: each discovery-stage rep turn contains at most one question seeking one answer, and the rep stops speaking after asking it.

A turn passes when it has one answer slot. A turn fails when it has two or more. The call-level result is the count of failed turns beside the count of discovery-stage turns, read per scenario over time. Keep it out of a blended skill score.

Load the row into your own call stages. If your playbook puts impact questions in stage two and decision process in stage three, the row applies to those stages by name. We hold that scoring changes behavior only when it is graded against your own methodology and call stages; a universal row cannot tell the rep which turns count as discovery in your process. For setting up and calibrating rows like this one across managers, see our rubric calibration guide.

The talk/listen ratio on a scorecard can be a useful input when you investigate. A rep who stacks often may hold the floor longer per turn, so a high rep share alongside many stacked turns is worth a look. Treat the ratio as a place to start digging, never as the pass bar. The pass bar here is the stack count.

A rubric row works when two managers can score it from the transcript alone, without having heard the call or knowing the rep.

How do you run the one-question drill?

Run it as a weekly 30-minute session: 5 minutes of setup, 15 minutes of drill, 10 minutes of debrief. These parameters are our recommendation, not research findings. Budget about 10 more minutes beforehand to pick a scenario from a live deal and choose the rubric row. That prep is real cost, and it belongs to the manager. Put the 30 minutes on the calendar as a recurring block separate from pipeline review; we hold that a coaching block survives only as a protected recurring calendar commitment.

Setup (5 minutes). State the single behavior: ask one question, then stop. Read the pass bar aloud.

Drill (15 minutes). The rep runs a discovery-stage scenario against a buyer persona, human or AI. The buyer rule matters: the persona answers only the easiest question in any stack and does not circle back. If the rep stacks, the transcript shows the cost directly, which is a gap the rep can see.

Pass bar. Zero stacked turns across the run's discovery-stage turns, and the rep gets an answer to each question asked before moving on. Both conditions come from the transcript.

What a failing attempt sounds like. The rep opens with "Walk me through how you handle this today, who owns it, and where it breaks down?" The buyer describes the process. Owner and breakdown never come up, and the rep says "Great, and what's your budget?" That one run contains a stack and an unrecovered skip.

Pick scenarios tied to real deal stages. Our discovery exit criteria scorecard lists the evidence a stage must produce, which tells you which questions the scenario must force the rep to ask one at a time.

A zero-stack pass bar and a buyer who answers only the easiest question make the cost of stacking audible immediately.

What does the debrief and re-run look like?

The debrief isolates one stacked moment, lets the rep find it first, and ends with a scheduled re-run. Do not cover pacing, tone, or the close in this session. We hold that a debrief changes behavior only when it isolates one moment and one behavior, so a debrief that lists three issues is a review.

Use this script. First, play or read one stacked turn from the transcript. Second, ask: "What did you ask there, and what did the buyer answer?" Wait. If the rep names the skipped question, the self-diagnosis is done. If not, point to the answer slots yourself, one at a time. Third, have the rep restate the turn as two separate turns aloud. Fourth, book the re-run before the rep leaves the room.

If one moment needs more time, we recommend a separate 15-minute 1:1 built around that single flagged moment. Hold it in addition to the pipeline 1:1, never in place of it.

Re-run rule. A failed drill is re-run on the same scenario within the week, with the same pass bar. Do not advance on a failed run, and do not accept "I get it now" as a substitute. Our position is that training counts only when it certifies observable behavior under pressure, and a rep's understanding is recall. Only a second transcript certifies the behavior.

Once the rep passes the original scenario, change the scenario and run it again. Check whether stacking returns when the buyer is hostile or the rep is nervous; a pass on an easy persona alone is thin evidence. For the debrief mechanics in more depth, see our one-behavior debrief guide.

A failed drill earns a re-run on the same scenario, because only a second passing transcript certifies that the behavior changed.

Where does the one-question rule bend?

The rule bends in three places, and a manager should name them before scoring so reps do not argue the rubric. First, a clarifying restatement is not a stack. "So the handoff breaks at review. Is that right?" confirms one thing and passes. Second, a two-part question with a single answer slot can pass: "Who signs off, and is it one person?" is arguably one slot, and the manager should decide it in advance and write it down.

Third, the rule is weakest when a buyer is rushed. A buyer with five minutes may want two quick questions in one turn. A rep who stacks deliberately, having agreed a short time-box with the buyer, is making a choice. The rubric scores the default, and the drill builds the default.

The trade-off is real. Single-question turns can feel slow, so put a short acknowledgment before each question instead of a bundle.

The drill also does not teach what to ask. It teaches the shape of the turn. Pair it with a content rubric from your own methodology and score the two separately, in separate debriefs.

Single-question turns are a default to build, not a law, and a rep should be able to break the default on purpose.