Anchoring bias is the tendency to rely disproportionately on the first number encountered in a decision or negotiation. All subsequent estimates are adjusted from that anchor, but adjustments are systematically insufficient. The anchor exerts influence even when it is known to be arbitrary. The antidote is to form an independent estimate before seeing any external number, and to go first in negotiations when you have done your research.
Where this came from
Amos Tversky and Daniel Kahneman documented anchoring in 1974 as part of their broader work on cognitive heuristics and biases. Their original experiment is one of the most striking in psychology. Participants watched a wheel of fortune spin to a number, either 10 or 65. They knew the spin was random. They were then asked: what percentage of African countries are members of the United Nations? Those who saw 65 guessed much higher than those who saw 10. An irrelevant, random number had measurably shifted their estimates.
Subsequent research confirmed the effect across domains where much higher stakes were involved. Strack and Mussweiler (1997) showed that anchoring affected judges' sentencing decisions: judges given a higher suggested sentence by prosecutors handed down longer sentences than judges given a lower suggestion, even when the judges believed they were reasoning independently. Northcraft and Neale (1987) showed that property agents' valuations were anchored to listing prices that were experimentally manipulated, and the agents, who were professionals, showed the same susceptibility as untrained participants.
The effect has been replicated consistently for more than 50 years across cultures, professions, and types of numerical judgement. Knowing about anchoring reduces its effect somewhat, but does not eliminate it. This is what makes it practically significant: awareness provides only partial protection.
How it works
The mechanism behind anchoring involves two processes working together. The first is that an initial number activates a selective search for information consistent with that number. If you hear that a house is worth Rs 2 crore, you start retrieving features that support that figure: the location, the renovation, the views. If you had heard Rs 1.2 crore, you would retrieve different features: the distance from the metro, the age of the building. The anchor changes which evidence you search for.
The second process is adjustment. Starting from the anchor, people adjust upward or downward based on other information. But adjustments stop as soon as the estimate enters what feels like a plausible range, not when they reach the objectively correct number. This means adjustments are nearly always insufficient. You land closer to the anchor than the facts warrant.
Anchoring is particularly powerful in three situations. First, when you are uncertain. Uncertainty makes any available reference point feel useful. Second, when the anchor comes from a credible source. A price stated by an expert feels more informative than a random spin of a wheel, even if neither is reliable. Third, when the negotiation is adversarial. In salary or commercial negotiations, anchors are often deliberately set by the other party, who understands their effect even if unconsciously.
The effect is not limited to numerical decisions. The first description you hear of a person, product, or situation creates an anchor for all subsequent evaluations. First impressions in hiring interviews, for example, are a form of anchoring: the initial impression set in the first few minutes influences the interpretation of everything that follows.
When to watch for it and when it matters most
Anchoring is most consequential in negotiation, financial valuation, project estimation, and performance appraisal. In salary negotiation, whoever names a number first sets the frame. In commercial negotiations, the opening offer is not a starting point for discussion; it is a deliberate anchor. In project budgeting, the first estimate produced, often in a preliminary scoping session, tends to survive into the final budget with insufficient adjustment for the risks uncovered in later planning.
Watch for anchoring when you are reviewing someone else's estimate before forming your own. Reading an analyst's report before making an investment decision, reviewing a supplier's quote before knowing the market rate, or seeing a consultant's proposed project budget before your team has formed an independent view are all contexts where anchoring can distort your judgement before you realise it is happening.
Adjustment Insufficiency
Even when people know an anchor is arbitrary, they adjust away from it far less than is rational. In experiments where participants are told explicitly that an anchor was generated randomly, they still adjust insufficiently. Knowing about anchoring reduces the effect by a modest amount but does not override it. This means the standard advice to "consider whether you are being anchored" is not enough. The practical correction must be structural: form your own estimate first, before encountering any external number, so that your anchor is your own independent judgement rather than someone else's opening position.
Run this framework on your actual decision.
DecisionsMatter.ai walks you through a structured 5-step analysis. Your first analysis is free.
How to apply it in practice
In any decision where you will be presented with an external number, form your own estimate first. Write it down. Then engage with the external information. Having a written prior estimate gives you a reference point that is your own rather than the other party's. This is the single most effective behavioural counter to anchoring.
In negotiations, prepare a clearly justified opening number before you enter the room. Your anchor should be ambitious enough to pull the negotiation in your favour, but grounded in data you can defend. An anchor that is so extreme it loses credibility can backfire; the other party stops treating you as a reasonable negotiating partner. The goal is an anchor that is stretching but defensible.
When you must evaluate an estimate you cannot avoid seeing first, run a deliberate counter-anchoring exercise. Ask yourself: if the first number I had seen were half as large, what would I now estimate? If it were twice as large, what would I estimate? The range those two scenarios generate tells you something about the legitimate uncertainty in the estimate, and it forces your mind to explore territory the anchor had implicitly placed off limits.
In project planning specifically, build budget and timeline estimates from the bottom up, task by task, before you see any top-line figure from a previous project or a stakeholder's expectation. Once you have your own bottom-up estimate, then compare it to external references. This sequence preserves your independent judgement and forces the anchoring conversation to be explicit rather than invisible.
This is one model from the upcoming Decisions Matter book.
30 mental models. 40 cognitive biases. A 5-step decision system. Get the relevant chapter in your inbox when it publishes.
Frequently asked questions
What is anchoring bias?
Anchoring bias is the tendency to rely heavily on the first piece of numerical or factual information encountered when making estimates or decisions. Once an anchor is set, all subsequent judgements are adjusted relative to that anchor, but the adjustments are typically insufficient. The anchor exerts a distorting pull on the final estimate even when the decision-maker knows the anchor is arbitrary or irrelevant. The effect was documented by Amos Tversky and Daniel Kahneman in 1974 and has since been replicated across legal, financial, medical, and negotiation contexts.
What are examples of anchoring bias in negotiation?
In salary negotiation, the first number stated sets the anchor for the entire conversation. A candidate who names Rs 20 lakh first will typically receive a higher offer than one who waits for the employer to go first, because the employer must negotiate from the candidate's anchor. In property sales, listing prices anchor buyers' willingness to pay, even when the listing price bears no relation to market value. In legal settings, studies have shown that the prosecutor's sentencing recommendation influences the judge's sentence, even when judges believe they are reasoning independently.
How do you counteract anchoring bias?
Generate your own independent estimate before receiving any external number. If you must evaluate a supplier's quote, decide in advance what you believe fair market value is before opening the proposal. If you are receiving a performance review, write down what you believe your contribution has been before hearing your manager's assessment. Creating your own anchor before encountering the external one reduces the distorting effect. When you cannot avoid an anchor, explicitly ask yourself: if the first number I heard had been twice as high or half as low, what would I estimate now?
Should you always make the first offer in salary negotiation?
The research generally supports making the first offer when you have enough information to set a favourable anchor. A well-researched opening number pulls the negotiation in your direction. However, if you lack reliable information about market rates, going first risks setting an anchor below what the other party would have offered, which can cost you significantly. The better rule is: make the first offer when you have done your research; let the other party go first when you are negotiating with incomplete information.