Somewhere in the last month, a number changed your behavior without asking permission. Maybe it was a “was $89, now $59” tag. Maybe it was a recruiter asking your current salary before naming the role’s budget. Maybe it was a real estate listing priced just high enough to make the next one look reasonable.

You didn’t choose that number. But your brain used it anyway — as the invisible ruler it measured everything afterward against.

This is anchoring bias, one of the most consistently replicated findings in behavioral psychology. It’s not a flaw that only affects careless people. It affects experts, professionals, and — in some of the most unsettling research on the topic — people who were told exactly how it works moments before it happened to them anyway. The bias doesn’t discriminate by intelligence or attention span; it discriminates by sequence, which is precisely what makes it so hard to train yourself out of.

Understanding anchoring won’t make you immune to it. But it will change which numbers you trust, and which ones you should treat as a starting position in a negotiation you didn’t know you’d entered.

The Science Behind Anchoring Bias

Anchoring bias was first formally documented in 1974 by psychologists Daniel Kahneman and Amos Tversky, in a study that had nothing to do with money at all. Participants watched a wheel of fortune, rigged to stop on either a low number or a high number, then were asked an unrelated question: what percentage of United Nations member countries were African nations? People who saw the low number guessed low. People who saw the high number guessed high. The wheel had no logical connection to the answer — and it moved people’s estimates anyway.

That single finding launched decades of replication across wildly different contexts. In one widely cited real estate study, researchers Gregory Northcraft and Margaret Neale gave professional appraisers a full property listing and walk-through, varying only the asking price shown beforehand. Appraisers who saw a higher listing price produced higher “independent” valuations than those who saw a lower one — and when asked afterward, most insisted the listing price hadn’t factored into their judgment at all. Separately, researchers studying legal decision-making found that judges and law students handed harsher sentencing recommendations after being anchored with a higher number from an entirely arbitrary source, such as a dice roll, in experiments specifically designed to strip out any legitimate legal reasoning from the number’s origin.

Dan Ariely’s version, detailed in *Predictably Irrational*, sharpened the mechanism further. He asked students to write down the last two digits of their social security number before bidding in an auction for ordinary items — wine, keyboards, design books. Students with high two-digit numbers bid sixty to one hundred twenty percent more than students with low numbers, for identical items, in the same room. The number had no relationship to value whatsoever. It was simply the first thing that entered their mind before the bidding began, yet it shaped every bid that followed as reliably as if it had been real market data.

What makes anchoring different from most cognitive biases is that it doesn’t require you to be uninformed. It requires almost nothing except sequence — whichever number you encounter first quietly becomes the reference point every subsequent number gets compared against.

How Anchoring Bias Affects Your Financial Decisions

Anchoring shows up most obviously at the checkout counter. A “was $129, now $79” tag doesn’t just tell you the current price — it hands you a comparison point you didn’t ask for. Even when the original price was invented, inflated, or never actually charged to a single customer, your brain treats it as real data and calculates a “savings” that may not exist. Retailers know this. In most jurisdictions, the practice is entirely legal.

Salary negotiations run on the exact same mechanism, with higher stakes. Whoever states a number first — a recruiter opening with a lowball figure, or a candidate blurting out their previous salary — sets an invisible ceiling or floor that the rest of the conversation tends to orbit around, regardless of what the role is actually worth in the market. Career coaches who tell candidates to “let the employer name a number first” aren’t being coy. They’re trying to keep their client from handing over the anchor for free — because once a number is spoken aloud, walking it back down or up rarely feels as natural as staying near it.

Real estate offers a third, slower version of the same trap. A home listed noticeably above comparable properties resets what “reasonable” looks like for every buyer who tours it — even buyers who’ve done their own research and believe they’re immune to marketing. And on the investing side, many people anchor to the price they originally paid for a stock, holding a losing position far longer than the fundamentals justify, simply because their purchase price feels like the “correct” reference point rather than an arbitrary moment in time.

Software and subscription pricing use a fourth, more deliberate version. A “Pro” tier placed directly beside a far more expensive “Enterprise” tier isn’t primarily trying to sell the Enterprise plan — it’s anchoring the Pro tier’s price down by comparison, so a genuinely expensive subscription starts to look like the sensible middle option. The Enterprise tier can go almost entirely unsold and still do its job, quietly justifying its own existence purely through the anchor it creates for the tier beside it.

The Wine Auction That Never Ends

One of the more striking extensions of Ariely’s original study came from follow-up research testing whether the anchoring effect fades with experience. It largely doesn’t. Participants who went through multiple rounds of anchored bidding continued adjusting their valuations around the same arbitrary reference points, round after round — suggesting the anchor doesn’t just distort a single decision, it can quietly recalibrate what someone considers a “normal” price going forward, compounding its effect over repeated purchases rather than wearing off.

This matters more in 2026 than it did when the original research was published. Dynamic pricing, personalized “was/now” displays, and algorithmically generated “recommended” price points mean the average person now encounters far more anchors per day, from more sources, than any experiment could easily replicate — each one a small, largely invisible nudge on the number that follows it.

Flight-booking sites are a clean modern example. The price you see on your first visit quietly becomes your personal anchor for that route, even if you never book that day — so when you return a week later and the fare has shifted, your sense of whether it’s “gone up” or “come down” is being measured against a number the algorithm effectively chose for you, not against any stable notion of what the flight is actually worth.

The Practical Reframe

Most advice about cognitive bias stops at “be aware of it,” which research on anchoring specifically shows doesn’t work. Ariely’s own subjects, told in detail how the SSN-anchoring effect functions, fell for a near-identical version of it minutes later. Awareness changes what you can explain afterward. It doesn’t change what happens in the moment.

The more useful reframe is this: treat the first number you encounter in any negotiation, sale, or high-stakes decision as a plant, not as information. A plant doesn’t need to be dismissed — it needs to be identified before it quietly becomes your reference point. Before responding to any opening number, ask a simple question: where did this figure come from, and does it have any actual relationship to the thing being valued? Often, the honest answer is no.

Experienced negotiators use a more active version of this same reframe: they anchor first, deliberately, precisely because whoever names a number first tends to control the range everything else gets discussed within. If you’re on the receiving end of an anchor, the countermove isn’t to quietly negotiate downward from their number — it’s to introduce a competing anchor of your own, grounded in your own research, before their figure has a chance to settle in as the default frame.

If this idea resonates, the full breakdown goes further.

The video covers two additional biases alongside anchoring — the relativity trap and the zero effect — and closes with the specific finding that explains why simply knowing about these patterns isn’t enough to stop them.

Conclusion

The uncomfortable part of anchoring research isn’t that people get manipulated by arbitrary numbers. It’s that expertise doesn’t fix it. Appraisers, judges, and negotiators — people whose entire job depends on independent judgment — get anchored just like everyone else, often while believing they haven’t been. Training and experience change how confidently people defend their number. They don’t reliably change whether an arbitrary anchor shaped it in the first place.

That should change how you treat your own confidence. The feeling of having made an independent, well-reasoned decision is not evidence that you did. Sometimes it’s just evidence that the first number arrived before you had a chance to ask where it came from — and by the time you’re evaluating the second number, the comparison has already been quietly decided for you.

The next time a number arrives before you’ve had a chance to think, that timing itself is worth noticing. Was it offered to inform you, or simply timed to arrive first — before you had anything to compare it against?