Never Split the Difference

The new rules: people are not rational

Chris Voss spent more than two decades at the FBI and became its lead international kidnapping negotiator. The book opens with Voss at Harvard, where he joins a negotiation course and holds his own against students and professors trained in the classic problem-solving model. That model, made famous by Getting to Yes, treats negotiation as a rational exercise: separate the people from the problem, focus on interests, invent options and use objective criteria.

Voss argues that this model misses the most important fact about people. They decide with emotion first and justify with logic later. He draws on Daniel Kahneman's work on fast, intuitive System 1 thinking and slow, deliberate System 2 thinking. Most decisions happen in System 1. So a negotiator who only argues facts is talking to the wrong part of the brain.

His alternative is tactical empathy: understanding the other side's feelings and mindset, and saying that understanding out loud. Empathy here is not sympathy or agreement. It is a tool for gathering information and lowering the other side's defenses. The goal of a negotiation, in Voss's view, is to discover as much as possible about the other side, because the facts you do not know yet often decide the outcome.

Listen first: mirrors and labels

Be a mirror and slow it down

The first tools are about listening. Voss says negotiators fail when they rush. They hurry to their own argument and miss what the other side is telling them. He advises you to slow down and make the other person feel safe.

Voice matters. Voss describes three tones:

• The late-night FM DJ voice: deep, calm and slow, with a downward inflection. Use it sparingly to state firm points without starting a fight.
• The positive or playful voice: the default. It signals that you are easygoing and relaxes both sides.
• The direct or assertive voice: rarely useful. It creates pushback.

The next tool is mirroring. You repeat the last one to three words, or the most important words, of what the other person just said, with an upward inflection. It sounds simple. It works because people tend to trust those who seem similar to them, and because a mirror invites the other person to keep talking and explain more. Voss describes a 1993 bank robbery in Brooklyn where mirrors helped the FBI learn key facts from a robber on the phone. He pairs mirrors with silence: say the mirror, then wait for at least four seconds.

Label the emotion, defuse the fear

Labeling means naming the other person's feelings out loud. You use phrases like "It seems like...", "It sounds like..." or "It looks like...". You do not say "I'm hearing that," because "I" puts the focus on you. After a label, stay quiet and let the other person respond.

Voss explains that labeling a negative emotion reduces its power. Naming fear or anger helps calm it. Labeling a positive emotion reinforces it.

He adds the accusation audit. Before the other side can attack, you list every negative thing they might think about you: "You probably think I'm being unfair. You may feel that I don't care about your situation." Saying the worst first takes the sting out. Usually the other side responds by saying it is not that bad.

Rethink "yes" and "no"

Beware "yes," master "no"

Most sales and negotiation training pushes for "yes." Voss argues that "yes" is often hollow. He separates three kinds:

• Counterfeit yes: said to make you go away.
• Confirmation yes: a simple reflexive affirmation.
• Commitment yes: a real agreement that leads to action.

People often feel trapped when pushed toward "yes." In contrast, "no" makes people feel safe and in control. So Voss recommends questions that invite a "no." Instead of "Do you have a few minutes to talk?", ask "Is now a bad time to talk?" Instead of "Do you agree?", ask "Is this a ridiculous idea?"

He gives a famous email tactic. When someone stops replying, send a one-line message: "Have you given up on this project?" The question invites a "no," and people often reply quickly to say they have not.

"That's right" beats "you're right"

The two words Voss most wants to hear are "that's right." They mean the other side feels truly understood. To get there, you summarize: you paraphrase their position and label their feelings until they agree that you have captured their view.

He contrasts this with "you're right," which people often say to end a conversation without changing their mind. Voss tells the story of a kidnapping in the Philippines involving the Abu Sayyaf group, where a summary of the captor's grievances finally produced "that's right" and a turning point. The aim is to make the other side feel heard before you ask for anything.

Bend their reality

Voss argues that compromise is often the worst outcome. If one party wants black shoes and the other wants brown, splitting the difference produces one black shoe and one brown shoe. Hence the title.

To avoid splitting the difference, he uses several tools based on how people actually perceive value:

• Deadlines are often arbitrary and more flexible than they seem. The person who feels deadline pressure is usually the one who concedes.
• The word "fair" is emotionally loaded. When someone says "I just want what's fair," they often mean you are unfair. Voss suggests answering with a calm offer to stop and review where you have treated them unfairly.
• Anchor their emotions with an accusation audit before an offer, so the offer looks better by comparison.
• Use loss aversion, from prospect theory: people will take more risk to avoid a loss than to gain an equal amount. Frame what they will lose by not acting.
• Let the other side go first on price when you can. When you must name a number, use a range whose low end is your real target, and use precise, non-round numbers, which seem more carefully calculated.
• Add non-monetary terms that matter to them but cost you little.

Calibrated questions and execution

Create the illusion of control

Calibrated questions are open-ended questions, usually starting with "what" or "how." They avoid "why," which sounds accusatory. Examples: "How am I supposed to do that?", "What about this is important to you?" and "How can we solve this problem?"

These questions give the other side the feeling of control while you steer the conversation. They also force the other side to work on your problem. Voss calls "How am I supposed to do that?" the most powerful form of no, because it says no without saying it. He tells a story from a Haiti kidnapping case where calibrated questions moved the kidnappers toward a release.

Guarantee execution

An agreement means nothing if it does not get carried out. Voss recommends "how" questions to get the other side to describe how they will deliver, such as "How will we know we are on track?" He also wants you to identify everyone who can block the deal behind the table, not only the person at it.

He gives tools for spotting fake commitment:

• The 7-38-55 percent rule, based on research by Albert Mehrabian, suggests that tone and body language carry much of a message's emotional meaning. Watch for mismatches between words and delivery.
• The Rule of Three: get the other side to agree to the same thing three times, in different ways. It is hard to fake consistent agreement three times.
• The Pinocchio effect: liars tend to use more words and more third-person pronouns.
• Use your own name. Voss calls this the "Chris discount": a person who feels they know you by name is more likely to treat you well.

Bargain hard and find the Black Swans

For price bargaining, Voss introduces the Ackerman model, taught to him by a former CIA officer:

1. Set your target price.
2. Make a first offer at 65 percent of the target.
3. Raise it in three shrinking steps: 85, 95 and 100 percent.
4. Use empathy and several ways of saying "no" between offers so the other side counters first.
5. Make the final number precise and non-round.
6. With the last offer, add a small non-monetary item to show you are at your limit.

He also describes three negotiator types: Analysts (methodical, data driven, dislike surprises), Accommodators (relationship driven, enjoy talking) and Assertives (time is money, want to be heard). You should adapt to the other side's type, not your own.

The last chapter is about Black Swans: unknown facts that, once found, change everything. Voss's company, the Black Swan Group, is named after the idea, which he credits to Nassim Nicholas Taleb. Every negotiation has a few. You find them by listening, asking questions and meeting face to face. Black Swans often create leverage, which Voss splits into three types: positive (what they want), negative (what they can lose) and normative (their own stated rules and values, which you can hold them to). He uses the 1993 Waco siege to show how failing to understand the other side's worldview can lead to disaster.

What the critics say

Most of the evidence is Voss's own stories from hostage cases, which some readers say do not always transfer to repeated business relationships. Some critics see the tactics as manipulative when used without real care for the other side. The 7-38-55 rule is also widely misapplied, because Mehrabian's studies were narrow and concerned feelings, not all communication.

Who should read it

This book suits anyone who negotiates salaries, contracts, rent, sales or family decisions and wants concrete phrases to use. It is especially useful for people who tend to argue facts and avoid emotions.