Negotiation: trade concessions, never just give

Most reps lose margin not because they negotiate badly, but because they hand out concessions for free. The buyer says 'it's too expensive,' the rep drops five percent, and nothing comes back. The problem is not the five percent itself. The problem is what you just taught the buyer: that pressure works and there is more to be had.

The prep happens before the call. Know where your floor sits and what you do if the deal collapses, which we covered in the piece on BATNA, and you already have the backbone to say no. But BATNA is the preparation. This is what happens once the buyer is across the table pulling at your price.

What does it mean to anchor with the first number?

Anchoring means the first number named becomes the reference point the whole negotiation moves around. It is a cognitive bias, not a negotiation gimmick, and it is stronger than most people assume. Tversky and Kahneman showed back in 1974 that a purely random number swayed people's estimates: a rigged wheel of fortune stopping on 10 versus 65 pulled median guesses to around 25 percent versus 45 percent, even though the number was obviously irrelevant. The brain adjusts from the first figure it sees, and adjusts too little. Put the first offer on the table and you set that number. Let the buyer do it and they anchor you instead.

This is not just a lab finding. Galinsky and Mussweiler found in 2001 that the first offer in a negotiation correlated with the final price at roughly r 0.85. Whoever opens drags the outcome toward themselves. For a seller that means you should name price early and name it with confidence, rather than waiting until the buyer has set their own low reference point.

Why talk price early instead of late?

Because timing decides whether you negotiate from strength or on the defensive. Gong analyzed 11,331 deals and found that reps who raised price on the first call won 42 percent of the time. Pushed to the second call, the win rate fell to 32 percent, to the third call 15 percent. Waiting in the hope of 'building value first' teaches the buyer that price is something sensitive to be tiptoed toward, and so it becomes the first thing they attack. Salesprep has a negotiation module where you practise against an AI buyer who pushes hard for a discount, and every call is scored on whether you actually got something back for each concession, so you see the trading reflex settle in over the weeks.

How do you trade a concession for something back?

Phrase it as a conditional if-then and say the condition out loud before you name the price. 'If you move up to three licences, then I can drop the setup fee.' Never 'I'll drop the setup fee' followed by silence. The difference is that the first version demands something in return, the second gives it away. Harvard's Program on Negotiation calls this making concessions strategically: every reduction should be conditional, not unilateral.

Label what you are doing, too. When you actually move, say it plainly: 'this is a big concession on my side, and I need something in return.' Labelling the concession makes it visible and harder to take for granted. A silent discount does not register as a sacrifice, it just registers as the new price.

What is logrolling and why doesn't it cut price?

Logrolling means trading across variables you each value differently, instead of fighting over price. The buyer may care most about payment terms, you may care most about contract length. So you trade: longer commitment for instalments. Neither of you gave on price, but both got something you valued more. Have the list of tradeable variables ready before the call: contract length, volume, payment terms, a reference case, a faster decision, onboarding scope. The more variables on the table, the fewer conversations are only about the number.

How do you defend the price without lowering it?

By removing the buyer's risk instead of your margin. Gong found that language that reverses risk, meaning guarantees, opt-out clauses and clear SLAs, raised win rates by roughly 32 percent. It works because a buyer's 'too expensive' is often really 'I'm not sure it delivers.' A 30-day opt-out clause rarely costs you anything if the product holds up, but it solves the buyer's actual worry without touching the price tag. Hand that over before you hand over a discount.

A concession with nothing asked back is not a negotiation. It's just a lower price you suggested yourself.

The reflex to trade rather than give is built through repetition, not by reading about it. Run a few negotiation roleplays a week in Salesprep, against an AI buyer who genuinely pushes, and watch the score for whether you got something back for each give. That curve moves fast.