Anchoring Bias

Anchoring Bias

Overview

Anchoring bias describes the strong human tendency to rely too heavily on the first piece of information received when making decisions. The first number, price, or estimate we hear becomes a mental reference point that influences all subsequent judgments.

Surprisingly, anchors work even when we know the number is random or irrelevant. In Kahneman and Tversky's classic experiment, participants who spun a wheel of fortune producing a random number were still influenced by that number when guessing the percentage of African nations in the UN. This shows anchoring bias is a fundamental aspect of how the human brain works.

Understanding anchoring bias is important because it affects nearly all financial and strategic decisions. Salary negotiations, product pricing, project estimates, investment valuations, all are vulnerable to the first anchor that appears. The party that controls the first anchor has a tremendous psychological advantage in negotiations.

Origin Story

Amos Tversky and Daniel Kahneman published their paper "Judgment under Uncertainty: Heuristics and Biases" in Science journal in 1974. This paper introduced three main heuristics that humans use when facing uncertainty: representativeness, availability, and anchoring-and-adjustment.

The most famous anchoring experiment is the "wheel of fortune" experiment. They spun a wheel that produced either 10 or 65 in front of participants. Afterward, participants were asked to guess the percentage of African nations that were UN members. The group that saw 10 guessed an average of 25%, while the group that saw 65 guessed 45%. The correct answer was around 30%.

Another surprising experiment involved multiplication. One group was asked to estimate the result of 8×7×6×5×4×3×2×1, while another group estimated 1×2×3×4×5×6×7×8. The first group gave a median estimate of 2,250, while the second group only 512. The correct answer is 40,320. Both groups drastically underestimated, but the first anchor from the number sequence influenced their estimates.

This research paved the way for modern behavioral economics. Kahneman received the Nobel Prize in Economics in 2002 for his contribution to understanding how humans actually make decisions in the real world, far from the rationality assumptions of classical economic theory. Findings about anchoring are now widely used in business negotiations, pricing strategy, and public policy.

The book Thinking Fast and Slow explains anchoring as a product of System 1, the part of the brain that works automatically and quickly. System 1 uses anchors as shortcuts to make estimates without deep analysis. System 2, which is slower and more analytical, should correct this bias but often fails because the adjustments made don't go far enough from the initial anchor.

Core Principles

1. The First Anchor Dominates Judgment

The first piece of information received has far greater psychological weight than subsequent information. The brain uses the anchor as a baseline, then makes adjustments that are almost always insufficient. This happens automatically at the System 1 level before our conscious reasoning activates.

In the context of negotiation, research shows that the first offer has a stronger correlation with the final outcome than subsequent counter-offers. Studies from the Program on Negotiation at Harvard Law School found that the first mover advantage in negotiation can reach 60-70% of the final settlement range.

Concrete example: In property negotiations, higher listing prices consistently produce higher selling prices, even when controlled for location and house condition. Identical houses listed $20,000 higher sell on average $15,000 higher, a 75% return on the anchor premium.

2. Adjustments from Anchor Are Always Insufficient

Kahneman and Tversky explain this as "anchoring-and-adjustment." People start from the anchor, assess whether it's too high or low, then adjust. The problem is, the adjustment almost always stops too early. This is because the brain seeks information that confirms the anchor (confirmation bias) and stops adjusting when it finds a number that "seems reasonable enough."

In experiments, they gave two groups different questions about Gandhi's age at death. The first group was asked "Was Gandhi more than 114 years old when he died?", the second group was asked "Was Gandhi more than 35 years old when he died?" Even though both numbers were clearly wrong (Gandhi died at 78), the first group gave significantly higher average estimates.

Practical implication: If you receive a salary offer of $70,000 when the market rate is $90,000, you tend to counter with $80,000-$82,000 and hesitate to go straight to $90,000. The $70,000 anchor pulls your estimate down even though you have market rate data.

3. Irrelevant Anchors Still Have Influence

The most surprising aspect of anchoring research is that clearly random or irrelevant anchors still influence judgment. This shows anchoring is a deep cognitive mechanism.

In the judge experiment mentioned by Kahneman, experienced judges were asked to determine punishment for a shoplifter. Beforehand, they were asked to roll dice that were manipulated to produce either 3 or 9. Judges who got 3 gave an average sentence of 5 months, while those who got 9 gave an average of 8 months. Yet they knew the dice were random and had nothing to do with the case.

This explains why marketing techniques like "was: $299, now: $199" are so effective even though customers know the "was" price might not be a real price that was ever charged.

4. Extreme Anchors Produce Extreme Estimates

The more extreme the anchor, the more extreme the final estimate, even when adjustment occurs. This is because adjustment stops at a point that "makes sense" relative to the anchor, falling short of the objectively correct point.

Application in pricing: Premium service providers often include packages or add-ons with very high prices. These items rarely sell, but their main function is anchoring. After seeing a $500 package, customers feel a $120 package is a reasonable middle ground choice, even though they previously only spent in the $50-$60 range.

In product pricing, SaaS companies place an "Enterprise" plan with a high price up front, even though their main target is the "Professional" plan. The extreme anchor makes Professional look reasonable and affordable.

5. Precise Anchors Are Stronger Than Round Numbers

Recent research shows that anchors with precise numbers (e.g., $37,850) have stronger influence than round numbers (e.g., $38,000). This is because precise numbers give the impression that detailed calculation has been done, while round numbers look like rough estimates.

In property negotiations, houses listed at $255,500 get higher offers than those listed at $256,000 or $255,000. Precision creates the illusion that the seller has specific justification for that number.

Strategy for negotiators: Use precise numbers for counter-offers. Instead of offering $85,000 for salary, offer $87,500. Precise numbers make the negotiation opponent feel you have data or calculations supporting that number.

Implementation Steps

1. Set Up Anchor Detection System: Before every important financial decision or negotiation, create a checklist with questions: "What's the first number I heard?", "Who mentioned this number and what's their interest?", "Do I have independent data to verify this number?" Save this checklist in a document or notes app that's easily accessible.

1. Research Independently Before Exposure to Anchor: If possible, gather benchmark data before you're exposed to numbers from the other party. For example, before a salary interview, research market rates from Glassdoor, Levels.fyi, or industry reports. Note credible ranges based on data, anchored in verifiable numbers drawn from research instead of gut feeling. This gives you an internal anchor based on objectivity.

1. Create an Extreme Counter-Anchor: If the negotiation opponent has already given the first anchor and you feel it's unfair, don't just adjust slightly. Create a new independent counter-anchor that's extreme enough to counter the bias. For example, if a vendor quotes a project at $150,000 and you think it's too high, don't immediately counter with $130,000. Research first, then counter with "$85,000 based on similar projects in this industry with scope X, Y, Z."

1. Use Ranges Instead of Single Numbers: When you have to mention the first number, consider giving a range instead of a single number. Ranges provide flexibility while still anchoring. Example: "For this project scope, the market rate ranges from $70,000-$95,000 depending on timeline and complexity. We can discuss where the appropriate positioning is." This range anchors your counterpart in the zone you want.

1. Document Historical Anchors and Outcomes: Create a simple spreadsheet or database that records: Decision type | Initial anchor | Objective data | Final decision | Actual outcome. Review quarterly to see patterns. Are you consistently influenced by anchors? How often is the initial anchor very different from objective data? This pattern helps calibrate your awareness.

1. Practice Delayed Response: When receiving the first anchor, train yourself not to immediately respond or adjust. Say "I need to review the data before responding." This delay gives System 2 time to activate and do proper analysis. Even a 5-10 minute delay can significantly reduce anchor influence.

1. Simulate Negotiations with Team: Before important negotiations, do role-plays where one person gives an extreme anchor. Practice how to respond without being anchored. Create several scenarios with different anchors. This practice builds mental resistance to anchor manipulation and increases confidence in giving data-based counter-anchors.

Brief Case Studies

Case 1: Fresh Graduate Salary Negotiation

Sarah, a fresh computer science graduate, received a job offer from a tech startup. HR offered $65,000 as the starting salary. Sarah had researched and knew the median for similar positions in that city was $82,000-$88,000 for fresh graduates with backgrounds from tier-1 universities.

Instead of immediately countering with a number in that range, Sarah was anchored by the $65,000. She felt asking for $85,000 would look too high and unreasonable. She countered with $75,000, hoping to meet in the middle around $70,000. HR countered with $68,000 and Sarah eventually accepted $72,000.

Sarah's friend, Michael, with an identical background, received an offer from a similar company. When HR asked about salary expectations, Michael (who had prepared) mentioned first: "Based on research from Levels.fyi and Glassdoor for fresh CS grads from our university, the range is $82,000-$92,000. I expect to be in that range." HR countered with $78,000, Michael countered with specific data about relevant skills and closed at $85,000.

This $13,000 per year difference is significant. Over 5 years, assuming 3% annual raises, Sarah will earn $100,000+ less than Michael just because she was anchored by the first offer.

Case 2: B2B SaaS Product Pricing Strategy

An analytics SaaS startup launched a new product. The pricing team initially targeted $199/month for the Professional plan as the core offering. They launched with a pricing page displaying three tiers horizontally: Starter $49, Professional $199, Enterprise $499.

The conversion rate to Professional was only 8%, with 78% of users choosing Starter and 14% going directly to Enterprise. The team felt Professional pricing was too high and considered lowering it to $149.

Before executing, they consulted with a pricing expert who suggested changing the order to maximize the anchor effect. The display was changed to vertical cards with the Enterprise package on the far left (the first position caught by the eye), followed by Professional and Starter. They also added a "Custom" tier with no price, just a "Contact us" button, and positioned it as the most premium option.

After the relaunch without changing any prices, the conversion rate to Professional rose to 23%, conversions to Enterprise rose to 18%, and Starter dropped to 59%. Revenue per customer increased 67% just by rearranging the display to control the first anchor.

They then added a new element: for the first two weeks, the price display started with a comparison "$999/month if buying separate tools: Analytics $299 + Dashboard $349 + API $351", then showed their bundle price. Conversions increased another 12% because the $999 anchor made $199 feel like a huge savings.

Case 3: Engineering Team Project Estimation

A tech company was planning a new feature for their mobile app. The Product Manager (PM) brought the requirements to an engineering team meeting and asked, "Can we launch in 6 weeks?"

The engineering lead, who hadn't had time to break down the tasks, felt pressured to answer quickly. Anchored by "6 weeks," they guessed, "Probably 8-10 weeks for MVP."

The team started working and after two weeks of detailed breakdown, they realized the scope was far larger than the initial estimate. There were integrations with three third-party APIs that hadn't been tested, UI complexity not visible in mockups, and performance optimizations that were mandatory. The real estimate became 16-18 weeks for a production-ready version.

The engineering lead felt trapped because they had already committed to "8-10 weeks" in front of stakeholders. They decided to cut corners and release a minimal MVP version at week 12, planning to patch technical debt later. The MVP launched with many bugs, customer churn increased 23% due to poor user experience, and the team spent the next six weeks firefighting issues.

Total time spent: 18 weeks to stabilize the product. If the engineering lead had rejected the anchor from the start and done a detail-based estimate before setting a timeline, they could have communicated realistic expectations of 16 weeks and delivered a high-quality product.

Lesson: In estimation processes, don't accept anchors from PMs or stakeholders before you have data. Responding with "Let me break this down first and come back with a data-based estimate" prevents anchoring bias and produces more reliable commitments.

When to Use and Avoid

Use understanding of anchoring bias in every negotiation situation, especially those involving numbers or prices. Be vigilant when you're the recipient of the first offer. Apply counter-strategies with independent research before being exposed to anchors. Strategically leverage anchoring when you have to present the first number: use data to justify your anchor and position it at an advantageous level.

Pay attention to anchoring bias when making forecasts or project estimates. Don't let timelines or budgets mentioned by stakeholders anchor your technical estimates. Always break down details before committing to numbers. Use historical data from similar projects as a more objective anchor.

Avoid giving anchors too early if you don't have enough information. In negotiations, if the opponent has more leverage or information, let them mention the first number then use data to counter. Don't anchor yourself by mentioning a number that's too low because you're afraid of rejection.

Situations where anchoring should be avoided: decisions that are purely technical or scientific where objectivity is critical. Don't use anchor manipulation in contexts where trust and transparency are important for long-term relationships. In medical decisions or safety assessments, always use objective data without anchor bias.

Practical Advice

Build a habit of "pause and research" before responding to any number. When someone mentions a price, timeline, or estimate, don't immediately react. Ask for time to verify with data. Framing: "That's interesting, let me check our benchmarks and get back to you." This pause gives System 2 space to evaluate objectively.

Create a decision journal that records anchors and outcomes. Every time you make a decision involving numbers, note: Initial anchor | Anchor source | Objective data you gathered | Final decision | Actual outcome. Review quarterly to identify patterns where you're vulnerable to being anchored.

In team settings, assign one person as the "anchor skeptic" whose job is specifically to challenge the first number that appears. Rotate this role so all team members practice questioning assumptions. Create a culture where asking "Where did this number come from?" is considered a healthy professional habit and part of the team's due diligence.

For pricing strategy, test various anchor points through A/B testing. Don't assume one pricing structure is optimal. Test different orderings, different comparison points, different precision levels. Measure conversion rate, total revenue, and customer lifetime value.

Train teams to give range estimates instead of point estimates, especially in project planning. Ranges provide flexibility and avoid false precision. Use the format "Optimistic: X, Most Likely: Y, Pessimistic: Z" to communicate uncertainty.

In high-stakes negotiations, prepare the anchor you'll use along with response scripts for various anchors the opponent might give. If they anchor at X, you already have counter-anchor Y with justification Z. This preparation reduces pressure to improvise and the chance of being influenced by bias.

Remember that awareness of anchoring bias alone isn't enough to eliminate its effect. Even experts who know about this bias are still influenced by it. What makes the difference is having systems and checklists to minimize impact. Invest time building systems to resist anchors.