- Published on
Book - Chris Voss: Never Split the Difference
I read Never Split the Difference by Chris Voss during an engineering management program. It's a great book on negotiation and not only big ones.
Chris Voss used to be a hostage negotiatior for the CIA, so his negotiations were quite high-stakes, but his methods can be used in regular conversations as well.
Why the book is good
The book is useful because Voss treats negotiation as information gathering under pressure.
That maps well to engineering management, because many bad decisions come from incomplete information:
- the deadline has a hidden commitment
- the design objection is really an ownership concern
- the conflict is about trust, not the ticket
Tactical empathy
Voss calls one of the core ideas tactical empathy. The name sounds a bit strange at first, but the concept is useful: understand the other person's view clearly enough to work with it.
Before pushing a decision, understand the pressure around it.
A stakeholder says:
We need this by Friday.
A weak version of the conversation starts arguing about the date right away. A better version tries to find the missing context:
It sounds like Friday is tied to something outside this team.
That kind of response does more than buy time. It invites the real constraint into the room: Maybe sales promised it, maybe there is a customer demo or maybe Friday is arbitrary. Each of those leads to a different conversation.
Mirroring
Mirroring is one of the smallest tools in the book. You repeat the important phrase back.
We need this by Friday.
By Friday?
It feels almost too small to matter, but people often clarify themselves when given a little more room.
That is familiar from product and architecture work. The first requirement is rarely complete. "We need a dashboard" might mean an export. "We need this to be real time" might mean every five minutes. "The system has to be scalable" might mean nobody knows the expected traffic yet.
Mirroring helps uncover the real requirement without turning the conversation into cross-examination. It also helps people feel heard.
Labeling
Labeling means naming what seems to be happening.
It seems like there is a trust issue around the migration plan. It looks like this decision feels already made. It looks like you are upset.
This can feel awkward the first few times. But tension often gets worse when everyone pretends it is not there. A decent label can lower the temperature and make the real issue discussable.
Calibrated questions
Voss uses calibrated questions to make the other side help solve the problem.
- What are we missing?
- How would we do this?
- What risk are we accepting if we choose this?
- What would we stop doing to make room for this?
The main idea is to start with "What" and "How", because they need thoughtful responses. Avoiding "Why" is useful, because it usually sounds accusatory ("Why would we do this?"). Instead you can ask ("How would we make this scale?"). And here you kind of need to keep an open mind. Maybe you are right and the other person will think about it and reach the conclusion that you are right ("Yeah, we could never make it scale. It was a bad idea"). Or maybe they have a piece of information that'll help you ("We'll just add caching and xyz") and they have addressed your concerns.
The main thing is that by asking, you are leaving them in control and letting the other person solve your problem. Voss used this in negotiations where he would say "How do I do that?", when they asked for super high amounts of ransom.
My take
The book is useful for better conversations. It's an entertaining read with a lot of FBI anecdotes, I would recommend reading it.