The Tactical Empathy Discovery Playbook: Voss's five tools for discovery and hard conversations

01 Open with an accusation audit. Say the worst thing first, out loud.

Before you pitch, ask, or propose, list the worst things the counterpart could be thinking about you, your product, or your motives. Then say those things first. Negative emotion keeps its force while it is unspoken. The counterpart rehearses the objection silently while you talk past it, and it ambushes the conversation later. Naming it first drains the charge and signals you are not hiding the obvious. Say the worst thing they could think about you, first, out loud, and watch the negative emotion drain

Calibrate the audit to the moment:

The audit has to be specific and accurate. A generic admission does nothing. If the worst thing you name is not what they were actually thinking, it reads as evasive and creates a concern that was not there. Skip it in long-trust enterprise relationships where it can re-introduce worries the buyer had moved past.

02 Mirror the last 1-3 words. The elaboration is where the deal is.

Once the conversation opens, listen for the final one to three words the counterpart emphasizes and repeat them back with a slight upward inflection. A mirror is the lowest-cost move that pulls information toward you. It imposes no direction, pushes back on nothing, and opens a small social vacuum the counterpart fills out of reflex. What they fill it with is almost always more valuable than what they said before. Mirror the last 1-3 words, silence forces the counterpart to elaborate, and the elaboration is where the deal is

• Counterpart: "We tried something similar last year and it didn't really stick."
• You: "Didn't really stick?"
• Counterpart: "Well, the rollout went fine, but procurement wouldn't sign the renewal because legal flagged the data flow."

The mirror surfaced the procurement and legal dynamic that a direct question would never have reached. Use it sparingly. Run too many in one call and it turns obvious and damages rapport. Deliver it as neutral curiosity, not skepticism, or it converts from invitation to challenge.

03 Label the emotion when it surfaces, before they have to defend it.

When the counterpart's tone shifts (frustration, hesitation, eagerness, fear) name it neutrally. "It sounds like you're worried about how the team will react to another tool." "It looks like the timeline is the part that doesn't work for you." A label acknowledges the emotion without conceding the position. "I see you're worried about budget" is not "I agree we should drop the price." That decoupling is what lets the conversation move from emotional rails to rational ones without either side feeling unheard. Label the emotion before they have to defend it, "it sounds like you're worried about..." disarms the room

Voss's own framing for why logic alone fails here:

"No one anywhere is teaching anyone that presenting a logical argument is an emotionally intelligent way to accomplish anything."

· Chris Voss, Never Split the Difference, 2016

"How am I supposed to do that?", give the other party the illusion of control and they solve your problem for you

Get the label wrong and it reads as presumptuous, which damages trust ("I'm not afraid of anything, I'm just asking"). In engineering-led B2B procurement and cultures that prize emotional reserve, labeling can add friction where the buyer wants facts only. Read the room.

04 Use calibrated questions to hand over the illusion of control.

Once rapport holds, deploy "How am I supposed to..." and "What about this is important to you?" questions. Instead of refusing a demand, you ask how to meet it, and the counterpart, now invited into joint problem-solving, usually concludes on their own that the demand needs to move. Voss reports the counterpart either modifies the ask, brainstorms alternatives, or reveals resources that make it feasible. "How am I supposed to do that?", give the other party the illusion of control and they solve your problem for you

• "How am I supposed to make a case for this if procurement won't sign?"
• "What about the rollout last year are you most worried about repeating?"

The questions reframe the conversation from positional ("I want X, you want Y") to cooperative ("how do we both get there"). Drill them so reps reach for them under pressure rather than improvising. They fail in heavily scripted procurement where the buyer cannot deviate, and they read as manipulative once a buyer recognizes the pattern, so do not over-run them.

05 Ask questions that earn a "no" when you need real engagement.

Conventional selling hunts for "yes." Voss inverts it: ask questions where "no" is the safe answer. "Is now a bad time to talk?" outperforms "Do you have a few minutes?" because the recipient asserts control without conceding anything. Re-engagement emails work the same way: "Have you given up on solving X?" often earns a defensive "no, but..." that reopens a dead thread. Ask questions that earn a "No", saying no makes people feel safe; being pushed for yes makes them defensive

"Saying 'no' makes people feel safe, secure, and in control, whereas being pushed for 'yes' makes them defensive."

· Chris Voss, Never Split the Difference, 2016

Reserve no-oriented questions for moments that matter. Used on trivial confirmations they turn into fishing, and in high-trust referral selling a direct opener ("I'm following up on the intro from X") beats the technique.

06 Listen for "That's right." It is the signal to propose, not to keep summarizing.

When you summarize the counterpart's situation accurately enough that they say "That's right" (not "Yes," which is often a stalling token) the conversation has crossed from positional to cooperative. They feel fully seen and there is nothing left to defend. That is the moment to propose. "That's right", not "yes", is the moment a negotiation actually shifts

• You: "So if I'm hearing it right, the timeline isn't the real problem. The team's residual frustration from last year's rollout is what makes any new tool feel risky."
• Counterpart: "That's right."

"That's right" emerges only after accurate labeling that names the why, not just the what. Deliver the summary as a statement, not a question. A hedged or generic summary earns "yes," not "that's right." Do not treat it as a checklist item ("did they say it yet?"). A faked moment is detectable, and the signal is hard to surface over email or Slack.

07 Hunt the Black Swans. Treat each call as a search for 3-5 facts that change everything.

Voss's claim is that every negotiation holds three to five hidden facts that, if discovered, would change the outcome, and they surface from rapport, not research. Usually they are authority dynamics, hidden constraints, internal politics, prior commitments, or unstated emotional stakes. None appear in public data. None get volunteered to a direct question. They leak through the elaboration after a mirror, the hedge after an accusation audit, the relief after a "that's right." Every negotiation has 3-5 hidden facts that change everything, they surface from rapport, not research

Build the hunt into the deal record:

• In the CRM, track an explicit field: what is the unknown that could change this deal? Hidden authority, internal politics, prior bad experience, regulatory constraint, budget timing.
• After each call, name what you learned that you did not prepare for. Those are your Black Swans.
• Account plans flag "what we don't know yet" next to "what we know." The unknowns are the work.

Reps who surface a Black Swan per major deal tend to close at higher rates than reps who over-rotate on prep. The discipline is sharpest in high-ACV, low-volume enterprise deals and structurally suppressed in transparent RFPs where the buyer discloses the criteria.

08 Diagnose indecision before you push. Hammering the status quo backfires.

Most stalled B2B deals are not lost to a competitor. They are lost to the buyer's fear of getting it wrong. Dixon's analysis of millions of recorded sales conversations found 40 to 60% of deals lost to "no decision" and roughly 87% of deals carrying medium-to-high indecision somewhere in the cycle. The dominant cause is fear of messing up, distinct from fear of missing out. 40-60% of B2B deals are lost to "no decision", and 87% of deals show medium-to-high indecision

The reflex under pressure is to make the cost of inaction bigger. That is the wrong move on an indecisive buyer. When reps sense hesitation, 73% double down on hammering the status quo, and it backfires 84% of the time. The buyer hears "the cost of getting this wrong is huge" and freezes harder. When buyers are indecisive, 73% of reps double down on hammering the status quo, and it backfires 84% of the time

So diagnose the mindset before you choose a tactic:

For the FOMU buyer, reach for JOLT: judge the level of indecision, offer your personal recommendation rather than a menu of options, limit the exploration so they do not research themselves into paralysis, and take risk off the table with a pilot or a guarantee. The biggest source of lost B2B deals is customer indecision, not competitor wins. JOLT them out. The tools from steps 1 through 7 are how you read which state the buyer is in before you act.

09 Develop the need with implication questions. The buyer's own voice is the argument.

In large sales, the seller behavior that separates star performers from average ones is concentrated in the investigation stage, not the close. Top and average reps differ mainly in the questions they ask, not in closing technique, presentation, or charm. Star performers are made in the investigation stage, not the close, top reps differ from average reps mainly in the questions they ask

The highest-payoff question type is the implication question, which asks the buyer to trace the second- and third-order consequences of their problem. Top performers ask roughly 4x more of them. Each step the buyer walks ("what happens if this persists, what does the team do to compensate, what does it block") amplifies the felt severity, and the buyer talking through the cost is more persuasive than any claim you make. By the end of a good sequence, the buyer has talked themselves into the explicit need. Top reps ask 4× more Implication questions, the highest-leverage question type in large sales

This is where the Voss stack and SPIN reinforce each other. Mirrors and labels keep the rapport conditions under which an implication question lands as genuine curiosity rather than a sales trap. Ask implication questions on a real problem from real interest. Ask them as a checklist and the buyer detects the technique.

10 Run the same tools as switch interviews for voice-of-customer research.

The five-tool stack handles the live conversation. Moesta's switch interviews extend the identical mechanisms across 8 to 15 buyer and non-buyer conversations to surface the verbatim resistance language that powers positioning. Trace a recent purchase backward through anxiety, struggle, and the decision moment. The buyer narrates the problem in their own words, not your category vocabulary, and the switch trigger (the moment they decided to look) tells you what campaign to fire and when. JTBD interviews surface the customer's actual language and the switch trigger

"JTBD interviews help you define a customer's language — what they actually mean by 'easy' — the root cause of why they switched, and how their story connects, all of which directly impact your product marketing."

· Bob Moesta, Intercom podcast, 2020

Use the Voss tools inside each interview: open with an accusation audit, mirror to invite elaboration, label the emotion, listen for "that's right." Then aggregate across interviews for the pattern. When the market itself breaks, the same instinct applies at the company level: switch from selling mode to listening mode and let the next product cycle come from what you hear. Switch from selling mode to listening mode when the market breaks

"We realized, we cannot be out there necessarily selling right now. What we need to be doing is to be much more receptive to what customers are saying, versus trying to sell them on our product. So we changed gears into listening mode to say, ‘Let's use this time to just go back and understand how people are using our product.’"

· Shruti Kapoor, Clari blog, 2022

A caution on the AI you use to process transcripts. "Verbatim" is too vague an instruction. A model will stitch sentences, trim hedges, and produce output that looks sourced but is fabricated. Lock the constraints: one uninterrupted customer turn only, no merging or cleanup, preserve the hedges and fillers, keep the speaker label and timestamp. AI quotes are only as good as the instruction that produced them.

11 Run the surfaced behavior through the incentive lens.

Rapport produces information. Incentives explain it. Once a conversation surfaces something that puzzles you (a champion who goes quiet, a procurement block that appears late, a stakeholder who keeps stalling) ask what is actually rewarded in that person's environment before reaching for a psychological or competitive story. People respond to incentives with near-mechanical reliability, and stated reasons describe how people should behave, not how they will. When behavior puzzles you, look at incentives, that's where every other model is downstream of

"Never, ever, think about something else when you should be thinking about the power of incentives."

· Charlie Munger, Poor Charlie's Almanack, 2005

Three questions resolve most of it. What is rewarded here? Who decides what gets rewarded? What does this person have to do to keep their job, status, or income? The puzzling behavior usually collapses into rational choice given the real payoff matrix.