Price Anchoring: The First Number Wins

On January 27, 2010, Steve Jobs walked onto a stage in San Francisco to introduce a product that didn't have a category yet. The iPad was too big to be a phone and too small to be a laptop, and every tech pundit in America had spent the prior month guessing what Apple would charge for it. The consensus landed around $999. That number had been floating through blogs, analyst notes, and cable news segments for weeks.

Jobs knew the number was in the room before he said a word.

Eight minutes into the pricing segment of the keynote, he put "$999" on the screen in huge type. He let it sit there. He talked about the engineering. He talked about the display. He talked about the software. Then he said, "I am thrilled to announce to you that the iPad pricing starts not at $999, but at just $499." The $999 slid off the screen and $499 replaced it, and the audience, a room full of journalists and analysts who had been trained to evaluate technology products for a living, applauded.

They applauded a price.

Not because $499 was cheap for a device nobody needed yet. Because $999 had been planted so deeply into their cognitive architecture over the preceding weeks, and then reinforced on the screen sixty seconds earlier, that $499 felt like a gift. The anchor had done its job. The first number established a reference point, and every subsequent number was evaluated in relation to it.

Price anchoring, the cognitive phenomenon where an initial number shapes all subsequent numerical judgments, is the most powerful tool in pricing strategy and one of the most reliably demonstrated effects in behavioral science. This post builds on the anchoring bias foundation with deeper neuroscience and broader business application, examining not just what anchoring does but why the brain is built this way, where the effect's limits actually fall, and how to deploy it in pricing architecture that serves both your business and your customer.

Why Is the Brain an Anchoring Machine?

The original anchoring experiment is so simple it's almost absurd. In 1974, Amos Tversky and Daniel Kahneman spun a rigged wheel of fortune in front of research participants. The wheel landed on either 10 or 65. Then they asked: "Is the percentage of African countries in the United Nations higher or lower than this number?" Followed by: "What is your actual estimate?"

A random spin of a wheel. Zero informational content. Participants who saw 65 estimated a median of 45 percent. Participants who saw 10 estimated 25 percent. A twenty-point swing caused by a number that everyone in the room knew was meaningless.

The finding launched a subfield. But it also raised a question that Tversky and Kahneman acknowledged would take decades to resolve: why does the brain do this?

The answer, developed through neuroimaging research in the 2000s and 2010s, involves two mechanisms operating simultaneously.

The first is what psychologists call "insufficient adjustment." When the brain encounters a numerical question, it doesn't compute the answer from scratch. It starts with whatever number is available, the anchor, and adjusts from there. The adjustment is consistently insufficient. The brain moves away from the anchor, but not far enough. Nick Epley and Thomas Gilovich at Cornell demonstrated in 2006 that this process is effortful. Adjustment requires cognitive resources. When people are distracted, tired, or rushed, the adjustment shrinks further and the anchor's influence grows.

The second mechanism is "selective accessibility," proposed by Thomas Mussweiler and Fritz Strack in 1999. When the brain encounters an anchor, it doesn't just use the number as a starting point. It selectively retrieves information from memory that is consistent with the anchor. Show someone a high price and the brain automatically activates reasons the product might be valuable. Show a low price and the brain activates reasons it might be cheap. The anchor doesn't just set the numerical starting point. It biases the entire information-retrieval process that follows.

Mussweiler and Strack tested this with a creative experiment. They anchored participants on whether Mahatma Gandhi died before or after age 9, or before or after age 140. Both anchors are absurd. No reasonable person would estimate Gandhi's age at death near either number. Yet participants anchored on 140 estimated significantly higher ages than those anchored on 9. Even implausible anchors activated different memory retrieval patterns, pulling different facts and associations into consciousness.

Functional neuroimaging has added a layer. A 2009 study by Brian Knutson and colleagues at Stanford used fMRI to observe the brain during pricing decisions. When participants saw a price before deciding whether to buy, the medial prefrontal cortex, which integrates value signals, showed activation patterns consistent with anchor-influenced processing. The insula, which signals "too expensive," activated relative to the anchored reference point, not to any objective standard of value. The brain's pain-of-paying circuitry was calibrated to the anchor. The number didn't just influence the decision. It reconfigured the neural architecture that made the decision.

This is why anchoring works even when people know about it. The mechanism isn't a conscious calculation that you can override with awareness. It operates at the level of memory retrieval and neural value computation. Gregory Northcraft and Margaret Neale's 1987 experiment with real estate agents, where professional appraisers were anchored by listing prices they claimed hadn't influenced them, demonstrated the point. Eighty-one percent of the agents denied the listing price had affected their judgment. Their appraisals showed it had.

Your customer's brain is an anchoring machine. It will use the first number it encounters to calibrate every number that follows. The only question is whether you provide that number or leave it to your competitor, your customer's last purchase, or a random figure they saw in a Google search.

How Did Williams-Sonoma Sell a $275 Bread Maker by Adding a $429 One?

In 1992, Williams-Sonoma was struggling to sell a home bread maker priced at $275. Sales were disappointing. The product sat on shelves. The conventional response would have been to lower the price, run a promotion, or discontinue it. Instead, the retailer introduced a second bread maker, a larger, slightly more featured model, at $429.

Sales of the $275 machine nearly doubled.

Nobody bought the $429 bread maker. That wasn't the point. The $429 model served a single function: it was the anchor. Before its introduction, the $275 machine sat in a psychological vacuum. The customer had no reference point within the product category to evaluate whether $275 was a reasonable price for a home bread maker. Is that a lot? Is it a little? The brain had no basis for comparison and defaulted to uncertainty, which, per Ellsberg's work on ambiguity aversion, produces inaction.

The $429 machine resolved the ambiguity. Now the $275 machine wasn't expensive. It was the affordable option. The customer wasn't deciding whether to buy a bread maker. They were deciding which bread maker to buy. And in that comparison, the $275 machine looked like the reasonable choice.

Itamar Simonson, a marketing professor at Stanford who studied the Williams-Sonoma case, formalized this as the "compromise effect." When consumers face three options, they disproportionately choose the middle one. The brain treats the middle option as the safe choice, the one that minimizes the risk of overspending and the risk of getting too little. A three-tier pricing page (basic, standard, premium) exploits this effect by design. The premium tier anchors high. The basic tier anchors low. The middle tier absorbs the majority of purchases, not because it's the best value but because the brain's risk-calibration machinery codes it as the safest bet.

Tiered pricing structures that follow this pattern are not manipulations. They're architecture. The brain needs reference points to make decisions. Without them, it stalls. With them, it computes. The question isn't whether to provide reference points but whether to provide them strategically or leave them to chance.

Where Does Anchoring Break Down?

Anchoring is among the most robust effects in behavioral science. It is not invincible.

In 2025, Ye Li and Carly Weigel published a study in Economic Inquiry that directly challenged the magnitude of anchoring effects in the pricing literature. They calculated the statistical power of major anchoring studies and found it routinely below 30 percent. When they replicated an anchoring study with adequate power, the reported effect shrank from 31 percent to 3.4 percent, a reduction by a factor of roughly nine. The directional finding held. The headline number did not.

This is the honest state of the science. Anchoring works. It works in the direction the literature claims. The size of the effect in real-world pricing contexts is almost certainly smaller than the laboratory demonstrations suggest, because those demonstrations were underpowered and subject to publication bias.

The practical boundaries are worth understanding.

First, anchoring is strongest when the customer has low knowledge about the product category. The real estate agents in Northcraft and Neale's study were influenced by listing prices, but less than the amateurs in the same study. Expertise reduces the effect without eliminating it. A customer who has bought enterprise software for twenty years will be less anchored by your pricing page than a first-time buyer. But "less anchored" is not "unanchored."

Second, anchoring weakens when the anchor is transparently implausible. Mussweiler and Strack's Gandhi experiments showed that even absurd anchors have some effect, but the magnitude decreases with implausibility. A SaaS company pricing its starter tier at $49/month and its enterprise tier at $49,000/month will raise eyebrows, not revenue. The anchor must be high enough to shift perception but plausible enough that the brain takes it seriously.

Third, anchoring is modulated by motivation and cognitive load. Epley and Gilovich showed that adjustment from the anchor requires effort. When people are motivated and have the cognitive resources to deliberate, they adjust more. When they're tired, distracted, or decision-fatigued, the anchor has more influence. This is why late-night infomercials, crowded trade show floors, and Black Friday sales environments produce larger anchoring effects. The customer's adjustment machinery is running on fumes.

For founders, the calibration lesson is this: anchoring is a reliable tool, not a magic trick. Use it where the customer lacks category expertise, where your high-end option is plausible, and where the decision context favors fast evaluation over deep deliberation. In those conditions, the first number wins.

How Do You Build an Anchoring Architecture into Your Pricing?

The most effective pricing pages in SaaS, e-commerce, and professional services are not lists of options. They're narrative structures that guide the brain through a sequence of anchors.

Consider Salesforce's pricing page. The enterprise tier appears first, at $500 per user per month. The eye reads it, the brain registers it, and every number that follows is evaluated in relation to it. By the time the customer reaches the $25 per user starter plan, that price feels almost trivially small. Five percent of the anchor. The brain's selective accessibility mechanism has spent the last thirty seconds retrieving reasons why CRM software is valuable and expensive, primed by the $500 anchor. The $25 tier benefits from all that activated value information while triggering almost none of the pain-of-paying response.

Apple's product lineup executes the same strategy with physical products. The iPhone Pro Max, starting at $1,199, anchors the category. The standard iPhone at $799 is evaluated not against the abstract question "is $799 a lot for a phone?" but against the concrete reference point of $1,199 for the premium model. Every feature comparison the customer makes is occurring in a cognitive environment where $1,199 is the ceiling. The $799 model isn't expensive. It's $400 less than the alternative.

The sequence matters. A 2012 study by Suk, Lee, and Lichtenstein found that presenting prices in descending order (high to low) increased purchase likelihood and average spending compared to ascending order. The first number in the sequence serves as the anchor, and a high first number activates value-consistent information retrieval. A low first number activates cheapness-consistent retrieval. The same products, presented in the same context, with the only difference being the order in which prices appear, produce different purchasing behavior.

The napkin version of pricing architecture is: anchor high, present the target in the middle, and let the low tier serve as a safety net. The premium tier isn't there to sell. It's there to make your real offering look reasonable.