Definitions
Position: what someone says they want. A stated outcome. The return value.
Interest: the underlying variables, constraints, and incentives that produced the position. The function body.
Two people can hold opposing positions and have compatible interests. Two people can hold identical positions and have incompatible interests. You cannot tell which situation you're in by looking at positions alone.
Negotiating at the position level is like debugging by reading only the return value.
Why Positions Produce Zero-Sum
A position is a single point in solution space. When two positions conflict, the only available move is compromise: split the distance, trade concessions, or walk.
This is zero-sum by definition. The pie is fixed. Every point you win, they lose.
Interests define a region in solution space. Multiple positions can satisfy the same interest. When you operate at the interest level, the number of possible solutions explodes. You're no longer dividing a fixed pie. You're redesigning its shape.
Position-Level
One solution per side
Compromise = loss for both
Zero-sum by structure
Win/lose or lose/lose
Interest-Level
Many solutions per interest
Creative structure = gain for both
Non-zero-sum by design
Expand before dividing
Concrete Example
Enterprise deal. Platform vendor negotiating with a large financial services customer.
POSITIONS (surface)
Customer: "We need a 40% discount."
Vendor: "Best we can do is 10%."
At this level, it's a pricing tug-of-war. Pure zero-sum. Someone loses.
INTERESTS (underneath)
Customer's actual interests:
Procurement needs to show cost savings to justify renewal sign-off
CISO needs SOC2 compliance by Q3 or faces audit risk
Engineering lead wants migration support, not a cheaper tool
Vendor's actual interests:
Protect ARR for quarterly reporting
Land a reference account in financial services
Reduce churn risk by deepening integration
Now the solution space opens. Multi-year commitment at modest discount (procurement wins, ARR protected). Compliance accelerator package included (CISO wins, vendor deepens integration). Dedicated migration engineer (eng lead wins, churn risk drops). Case study agreement (vendor gets reference).
Same deal. Completely different structure. No one split the difference on price.
The Hidden Drivers
Interests are not rational preferences. They're outputs of messy human systems.
Risk โ what happens to them personally if this goes wrong
Incentives โ what they're measured on, what gets them promoted
Time pressure โ fiscal year end, board meeting, audit deadline, competitor move
Politics โ who else is watching, who gets credit, who gets blamed
Personal gain โ career trajectory, reputation, relationship leverage
The stated position almost never reveals which of these is dominant. That's the work.
Moving from Position to Interest
A protocol. Not a trick.
1. Hear the position. Don't react. Don't counter. Register it.
2. Ask what it solves. "If we did exactly that, what does it unlock for you?" (reveals the outcome behind the ask)
3. Ask what drives the timeline. "What's making this urgent now?" (reveals external constraints)
4. Ask who else cares. "Who else is involved in this decision?" (reveals political structure)
5. Ask what failure looks like. "What happens if we don't get this done?" (reveals risk and stakes)
6. Reframe. "So the real problem is X, and the discount was one way to solve it. Can we look at other shapes?"
Step 6 is where the transition happens. You're no longer responding to their position. You're co-solving their interest. That's a fundamentally different conversation.
Failure Modes
Assuming Interests
You project your model onto them. You solve a problem they don't have. You feel clever. They feel unheard.
Confusing Stated with Real
They say "we need flexibility." You hear features. The real interest is contract escape clauses because they got burned before.
Skipping the Person
You map the org's interest but miss the individual's. They need a win that's visible to their boss. The deal structure ignores that. It stalls.
When Selling Becomes Service
There's a moment in every real deal where the dynamic shifts.
You stop trying to get them to buy what you have. You start trying to solve what they actually need. Sometimes the answer is your product. Sometimes it's a different structure. Sometimes it's telling them they don't need you yet.
That moment is when you've fully moved from positions to interests. You're no longer optimising your outcome against theirs. You're optimising the system that contains both.
That's not altruism. It's alignment. And aligned systems produce better outcomes for everyone in them.
The Alignment Tree is a tool for doing this systematically. The discipline of delayed collapse is what keeps you in the interest layer long enough for it to work.