"Would one twenty work?" "No. A hundred." "One fifteen, that's the floor." Two people sliding along a single number line: closing means cutting into yourself, and not closing leaves both sides resentful. Worse, you have no idea where that hundred came from. A guess, an approval threshold, a line from their boss — three different problems with three different solutions, and you tried none of them.
"A hundred, noted. Can I ask how that number was set? Is that all that's left in this line for the year, or does anything above a hundred go through a different process?"
(Them: above a hundred it goes to group review, three months minimum.)
—— One question, and the problem stops being "cut thirty" and becomes "stay under that approval line." At least three solutions exist: sign 98 for the core scope this year and order the rest early next; split it into two purchases; or set up an annual framework agreement. You haven't moved a cent and their real constraint is solved.
—— And if the answer is "there simply is no money," you've saved three rounds of haggling and can go straight to cutting scope.
Mechanism: a position is a solution; an interest is a constraint. Conceding inside the solution space can only move value from one side to the other. Go back to the constraints and there are usually several feasible solutions, some of them better for both. People don't go back because they assume the other side's interests are the mirror image of their own — an assumption that is often wrong and, unasked, never gets overturned.
Boundary: not every position hides a reconcilable interest. Sometimes their interest genuinely is "spend less this quarter," and digging still leaves you at zero-sum. Sometimes the real interest is political (they need to show their boss they got you down), in which case give them something to take back as a win rather than more reasoning. And ask the "how" neutrally — said hard, it becomes an interrogation.
· What is my BATNA? Be specific: not "find another client" but "client A's 100k order needs an answer this week." A vague alternative is no alternative.
· My reservation price: derived from the BATNA, plus switching costs and risk. Write it on paper and take it in with you.
· What is their BATNA worth? If they don't buy from you, what do they do and what does it cost? That number is where every value argument lands.
· My target (not my floor — the number I actually want) and the reasoning behind it. Walk in with only a floor and you'll close near the floor.
· The highest-value thing to do in the 48 hours before a negotiation isn't rehearsing lines — it's getting one more alternative: one more quote, one more offer, one more supplier. It beats any script.
"I have two other offers, both higher." If it isn't true the risk is severe: it invites specifics (which company? what role? by when?), and once it collapses everything else you said gets discounted. An invented BATNA also can't hold you up: they say "you should think seriously about that one," and you discover you were never prepared to leave.
"I understand there's a ceiling on the level. I am in a final round elsewhere, so I want to make this decision properly. If base can't move, could we look at three things: a signing bonus, the vesting schedule, or an early review at six months? Any one of them makes this straightforward for me."
—— A real alternative steadies your voice without you acting, and splitting one axis into three gives them a way to pay you that doesn't break their rules. With no real alternative you need this step even more: you're not fighting for leverage, you're making it easier for them to say yes.
Mechanism: the only reason a deal exists is that agreement beats no agreement. So your floor isn't "what I want," it's "what I get if we don't agree" — an external number that their pressure cannot touch. It's also why threats and stalling barely register with someone who has a real alternative: their floor isn't at this table.
Three boundaries. One, the BATNA must be real: a fabricated one is checkable, and being caught costs you the credibility you already had. Two, don't brandish it — "I have other options," said with weight, is a threat and invites retaliation; simply say you're comparing. Three, a weak BATNA isn't a lost cause: the right response is to add issues (next section) and start improving your alternatives now, not to grind it out at the table.
· Add issues (before you quote): "Besides price, a few things usually shape this decision: timing, payment terms, scope, contract length. How do those rank on your side?"
· Ask for the ranking, not the ask: "If you could only protect one of these, which would it be?" A ranking is more honest than a demand, because it forces a trade-off.
· Trade in packages, never issue by issue: "If it's half up front and a two-year term, I can hit the December date and the price holds." Settling one issue at a time turns every issue into its own tug of war; only packages create trades.
· Put two or three equivalent packages on the table at once (MESO): "A: December delivery, 50% up front, 130k. B: January delivery, 60-day terms, 130k. C: December delivery, 60-day terms, 145k. Which do you lean toward?" Their pick tells you their weights, and since the three are worth about the same to you, any answer is fine.
"110 doesn't work, I can do 125." You've just conceded that this whole negotiation is a price axis. Whether you land on 125 or 120, it's only a question of how much you lose — and what they learned is: push and it moves.
"110 depends on what it comes with. Here are two combinations that both work for me: A, 110, drop the data migration, delivery mid-January. B, 130, full scope, December delivery, half up front.
—— Which matters more: saving the twenty, or having something to demo in December?"
(They pick B.) —— The moment they choose, you know their real constraint is the annual meeting, not the money. That information never surfaces in a pure price fight.
Mechanism: on a single issue, distribution is necessarily one side's gain and the other's loss. With two or more issues, any difference in marginal weights creates a trade that leaves both sides better off — a Pareto improvement. "Grow the pie" isn't a slogan, it's a structure you can compute, provided you know their ranking, which you only learn by asking and testing.
Boundary: some negotiations really are single-issue — a used phone off a listing, a one-off purchase of a standard part. Manufacturing issues there just wastes time and reads as slippery. The test is simple: besides money, is there a second thing that affects this decision? If not, talk price. Also, stating your priorities first can be exploited, so the safer route is to express priorities through whole packages rather than admitting out loud which issue you don't care about.
· My interests (three, ranked) | my guess at their interests (three, marked as guesses)
· My BATNA / reservation price / target | what their BATNA is worth
· Issue list: beyond price — timing, scope, payment terms, contract length, named case study, ongoing support — and mark which ones are cheap for me and valuable to them. That's your ammunition.
· Two or three equivalent packages, plus my opening question (usually "besides price, what else is shaping this decision for you?").
· Walk-away triggers: written, e.g. "below 120," "unlimited revisions," "payment beyond 90 days." Written down is what makes them hold — in the room, willpower loses to the urge to end the conversation.
"Fine, it's yours." You've taught them one thing: a late add-on works. It will come back at renewal and at every change request — and the three free months become the baseline they think they were always entitled to.
"I want this to happen. Support has a real cost, though, so I can't hand it over. Two ways: add 8k for a three-month support pack, or introduce me to one company like yours and I'll include the three months. Which is easier for you?"
—— Friendly throughout, not a step conceded, and two face-saving exits offered. Concessions always trade for something — that part never bends. How you decline can be very soft.
Mechanism: fairness is a motivational system independent of money. When an offer is judged unfair, people experience something close to disgust and will pay a real cost to reject it — even when rejecting leaves them with nothing. So "squeeze them to the floor" wins on arithmetic and frequently backfires on humans.
Boundaries. One, "soft on the person" is not a polite word for conceding: warm and unmoved is the complete form; warm alone gets eaten. Two, using "we're friends" to extract terms they wouldn't otherwise give is manipulation — you're spending the relationship, and it usually only spends once. Three, walking away is a decision, not a pressure tactic: saying "I can't do that one; if things change, come back to me" brings more returning business than a cold exit — and a bluffed walk-out, once seen through, means nobody believes any floor you name again.
Pick one negotiation you've been dragging out — a quote, a renewal, a raise, a lease. Today is preparation only. Don't open your mouth.