Never give a concession without getting something back. Lock the signing date and every ask first, then answer once with two or three options, each concession tied to volume, early payment, a longer term or a date.
The deal is in procurement, and every message asks for one more thing: a bigger discount, better payment terms, a rate lock.
Never give anything without getting something back. The whole mechanism is one word: not "we can do 10% off" but "we can look at 10% off if you commit to two years". Before any number moves, fix the signature date and get the buyer to name everything they want at once, price, billing and terms together, then answer once, with two or three priced options you would be happy to sign. Tie any concession to that date, and say at the start that if the date slips the concession is back on the table, so pulling it later does not look like a trick. Each concession you do make is smaller than the last and carries a heavier condition, and the things that cost you little but matter to the buyer go before anything that touches price.
What a discount can honestly buy, in Todd Caponi's list: a bigger volume, faster payment, a longer commitment, and a signature on a date you can forecast.
Armand Farrokh's one cut, from the 30MPC newsletter of May 2024. First he locks the timeline: before he takes anything to his CFO, he asks whether the buyer is "in a position to sign by [DATE]?", then gets agreement on every step before signature, the executive approvals, legal review and security review. He warns at that moment that if the CFO approves and the deal crosses into next month, the CFO will want to revisit it. Then he forces the full ask: what number gets it done, and what else do they want, because a rock-bottom price leaves nothing to trade on billing. "We give all of our proposals in one cut," he writes. He takes value out in exchange for price and asks for something back every time: a rate lock buys a multi-year term, a volume discount buys more seats. His closing pair of options: a $50,000 one-year deal with no rate lock, or a $90,000 multi-year deal with the rate lock.
Todd Caponi's four levers, in a room of five procurement leaders. On a $7.5 million deal with an oil services company, procurement opened with "In order to get this deal done, we will need a 35% discount." Caponi pointed to the four levers he had written on the whiteboard, volume, timing of cash, length of commitment and timing of the deal, and suggested they "go through them one by one" to see how close he could get them. They were not ready to buy for more divisions. For paying early he offered 5% off for prepaying year two and 10% for prepaying years two and three, which they took. He offered 5% more for each year beyond three, and they stayed at three. He paid them for committing to a signing date. His Four Levers post sets out the method, and his book The Transparency Sale argues that fake deadlines and late discounts teach buyers to wait.
RED BEAR's swap. "We can do 10% off" becomes an offer to explore the same 10% if the buyer commits to a two-year term. When the other side demands and offers nothing, the reply asks what they can commit to if you move. Its six guidelines run: give one, get one; concede only if you have to; get them to move first; how you concede matters more than what; cheap items first; slow and reluctant beats quick and eager. RED BEAR sells negotiation training.
Maddy Jackson's gives and gets across the whole deal. Early, a rate card buys the evaluation criteria and the names of the decision makers. Mid-deal, a proposal buys access to the person who signs. Late, a discount buys a fast, predictable signature date and case study rights, and your legal team's time buys a commitment on theirs.
Other trades that do not touch the price. Ian Koniak splits the payment: the annual price stays and billing comes in two halves six months apart. And for the procurement question "How am I supposed to do that?", which Chris Voss says procurement teams now use on sellers, he suggests treating it as a real question and answering with how it would work, then staying quiet.
Nobody has measured the rule. No study compares sellers who attached a condition to every concession with sellers who gave the same ground away, or sellers who answered in one cut with sellers who conceded in rounds. Caponi's deal is one case, told by the person who won it. The nearest experiment is forty years old: Stephen Clopton put practising industrial buyers through a negotiation in 1984 and found their savings depended on how the seller conceded and what the seller told them; the abstract does not say whether conditions were attached. Gong reports from 11,331 opportunities that win rates were about 10% higher when price came up on the first call, which is about when price enters a deal, not how the closing round runs. Gong sells the software behind it.
Shrinking concessions work, for the person doing the shrinking. Kian Siong Tey and colleagues ran seven studies with 2,311 people: offers that shrink, such as 1,500 then 1,210 then 1,180 then 1,170, made the other side think the limit was near, so they countered less ambitiously and got worse deals. The effect disappeared when the other side set their own target before starting.
Holding back too much ends in no deal. Marc Mertes, Dana Kunz and Joachim Hüffmeier found in 2023 that high demands with small concessions raised the rate of no agreement, driven by the other side's anger at getting too little value.
A choice of offers wins more, if the offers are ambitious. Geoffrey Leonardelli and colleagues ran six experiments in which one side opened with three offers of equal value to them instead of one package. The three-offer side claimed more value and was seen as more cooperative, but only when the offers were aggressive; with moderate offers the advantage disappeared.
Reciprocity is real but weak. Robert Cialdini and colleagues found in 1975 that 50% agreed to a small request after refusing a large one, against 17% asked only the small one, and Oliver Genschow's team repeated it in Cologne with 51% against 30% and 38%. Daniel O'Keefe and Scott Hale's review found the effect small overall and no link between how much was conceded and how well it worked. A bigger concession does not buy more goodwill.
Deadlines. The evidence that revealing a real deadline speeds the other side up, and that one too tight to work in ends in no agreement, is under Find the real compelling event.
The experiments are students and online panels in one-off bargaining over a fixed pie, and none compared a concession with a condition attached against the same concession given away. Nobody has tested the one-cut sequence in any setting. The one measured cost is that holding too much back raises the chance of no deal.
Lock the signature date and every ask before you move a number, then answer once with two or three options you would sign, each concession attached to something real: volume, early payment, a longer term, or a date you can forecast. Make each move smaller than the last. Say once, early, that a slipped date reopens the terms, and never invent an expiry.