Negotiation is not a fight. It's a search for the price both sides prefer to no deal.

Anchoring is the effect by which the first number named in a negotiation becomes the reference point against which every later number is judged. Once a figure is on the table, both sides tend to adjust from it rather than reason from scratch, so a high opening number pulls the eventual price up and a low one pulls it down. This is not a trick of confidence or charisma. It works even on people who know the anchor is arbitrary, which is what makes it worth understanding properly.

Where the effect comes from

The founding study is Galinsky and Mussweiler (2001), who showed that negotiators who made the first offer ended up with better outcomes than those who waited, largely because the first number set the frame for what counted as a reasonable counter. The site's own review of the wider evidence base, including how well the original finding has held up, is covered in Anchoring in Negotiation: What the 2023 Meta-Analysis Actually Shows. The short version is that the effect is real but not universal, and its size depends heavily on how much the other side already knows.

When anchoring first is the right move

Anchoring first helps when your number is defensible on its own terms and the other side has not yet formed a reference point of their own. A consultancy quoting a project fee before the client has seen comparable pricing is in a strong position to set the frame, provided the number is built from scope. Anchor before the buyer has anchored themselves.

Where it backfires

An anchor that the other side can immediately see through does more damage than no anchor at all, because it signals that your numbers are not to be trusted for the rest of the conversation. This is worse when the other side has more information than you do, such as a procurement function that has run this category of tender before and knows the market rate; an implausible anchor in that room reads as either naive or dishonest. It is also worth naming a boundary here: anchoring with a genuine number based on scope is ordinary negotiation, but anchoring with a fabricated reference price, an invented "list price" that was never charged to anyone, or a false comparison to a competitor's rate, is misrepresentation under the Business Protection from Misleading Marketing Regulations 2008. The tactic and the fabrication are different things, and only one of them is negotiation.

A worked example

An agency pitching a retainer knows the work will run to roughly 40 days a year at its normal day rate. Naming that figure early, before the client has stated a budget, sets a defensible anchor because it is traceable to scope. Naming a figure well above that, on the assumption the client will negotiate down anyway, risks the client checking the maths, finding the gap, and re-reading every other number in the proposal with the same suspicion.

The BATNA is the number you'd accept if this deal fell through

Your BATNA, Best Alternative to a Negotiated Agreement, is what you would do if this negotiation collapsed entirely: another client, another supplier, no deal at all. The term comes from Fisher and Ury's Getting to Yes (Harvard Negotiation Project, 1981), and it is the one number that should govern every concession you make in the room. A consultancy with two live proposals in play has a stronger BATNA than one chasing its only prospect this quarter, and that difference should show up in how you negotiate, not just in the spreadsheet behind it.

Work it out before you sit down, not while the buyer is talking. Write down the actual alternative: the fee from the other client, the cost of running this pitch again next quarter, the value of simply keeping the capacity free. Compare that figure to whatever price ends up on the table and you have your walk-away point. Negotiators who calculate this in advance concede less, because they know precisely what they are protecting; negotiators who guess at it in the room tend to fold at the first sign of resistance, because they have nothing firm to measure the offer against.

A concession that costs you nothing teaches the buyer to ask again

Every concession should extract something in return, because a free one simply moves the anchor and invites the next request. If a buyer asks for ten per cent off, the answer is not a straight discount. It is a discount in exchange for a shorter payment term, a reduced scope, or rights to use the work as a case study: something that costs you less than the ten per cent but is worth something to you. Give ground for nothing and you tell the buyer your price was never firm, and buyers who learn that keep pushing, because the last request worked.

This is where procurement's standard opening, "what's your best price", turns into a test. The response worth giving is a price attached to a condition.

The last five per cent usually costs more than it wins

The final stretch of a negotiation, the point where both sides have essentially agreed and one party pushes for a little more anyway, is where deals get damaged. Negotiation trainers sometimes call this the nibble: a small extra request made after the main terms are settled, timed for the moment the other side most wants the thing signed and is least willing to reopen it. It works in the short term because reopening the whole agreement feels disproportionate to the size of the ask.

It also tends to be remembered longer than the fee itself. A buyer who feels nibbled at the last moment carries that into the relationship, and it can matter more at the point of asking for the decision than the discount itself ever did. The two or three per cent squeezed out at signature is rarely worth what it costs in goodwill during delivery, renewal or the next negotiation. Knowing your BATNA lets you recognise this moment for what it is and decide, calmly, whether the last five per cent is worth losing the account over.

Most of what gets quoted as "negotiation research" in sales training comes from a narrow slice of psychology: laboratory experiments, often using students or MBA candidates, negotiating for course credit or a small cash prize. It is a reason to read them carefully before applying them to a seven-figure deal where the other side has spent months building a business case internally.

Where the classic findings come from

Loss aversion, the tendency to weight a loss more heavily than an equivalent gain, comes from Kahneman and Tversky (1979), working with hypothetical choice scenarios. The core effect has been replicated many times since, in different forms and different populations, which is why it holds up as a working assumption: a buyer will usually fight harder to avoid losing a discount they already have than to gain one they do not yet have. The size of the effect, and how cleanly it transfers to a commercial negotiation over price and scope, is less settled than the popular version of the finding suggests.

Anchoring is the clearest case of research meeting practice

Anchoring, the idea that the first number named sets the reference point for everything that follows, is one of the more heavily studied effects in negotiation research, and also one of the more contested when it comes to how strongly and how reliably it applies outside the lab. A 2023 meta-analysis pooling the anchoring literature is covered in detail on this site, including which conditions the effect held up under and which it did not: see Anchoring in Negotiation: What the 2023 Meta-Analysis Actually Shows. It is worth reading before deciding whether to anchor first on a specific deal, because the effect is not unconditional.

Replication is not a technicality

A study that has not been independently reproduced is a hypothesis with good manners, and a lot of the advice circulating in sales training rests on a single study from decades ago that nobody has repeated since. The same question, whether a popular effect survives when other researchers try to reproduce it under different conditions, comes up outside negotiation too: see Social Proof in B2B Sales: What Replicated and What Did Not for a comparable case in an adjacent area of persuasion research. The practical habit worth building is asking, before adopting any technique for a high-value negotiation, whether it rests on one study or many, and whether those studies looked anything like the room you are about to sit in.

Ongoing coverage of negotiation and persuasion research, updated as new studies and meta-analyses appear, sits on Fresh Research on Persuasion, Negotiation and Selling.