Most organizations expect their important decisions to be guided by data, expertise and careful analysis. However, in practice, that is only part of the story. Even well-run teams carry assumptions, mental shortcuts and blind spots with them, usually without realizing it.
In low-stakes situations, those habits are mostly harmless. But for organizations managing major investments, public resources or complex trade-offs, they can quickly become expensive.
This article looks at ten of the most common decision-making biases afflicting organizations, the situations that bring them about and practical ways to reduce their influence through better decision-making process design. This emphasis on process design is because that’s where biases can be reduced most effectively.
What are decision-making biases?
Decision-making biases are systematic patterns in how people interpret information and evaluate choices. Rooted in the psychology of judgment first analyzed by Nobel laureate Daniel Kahneman and Amos Tversky in the 1970s, rather than being random errors, biases are predictable and recurring, especially when decisions are made under conditions of uncertainty, complexity and time pressure.
In the workplace, these biases shape how options are framed, what evidence is emphasized, how risks are weighed and how final decisions are reached and justified. Because most organizational decisions are made in groups, individual biases interact with team dynamics – which makes them even harder to detect and more likely to compound.
What are the 10 most common decision-making biases in organizations?
The ten biases summarized below are the most potentially important for organizations making complex, high-stakes decisions. Each bias is explained in turn below with a workplace example and a practical, process-based solution.
- Authority bias – when hierarchy outweighs evidence
- Groupthink – when conformity silences doubts
- Sunk cost fallacy – when already-incurred costs drive future decisions
- Anchoring bias – when the first information received sets the tone
- Availability heuristic – when recent events override long-term priorities
- Confirmation bias – when evaluation becomes justification
- Choice overload – when having too many options overloads choices
- The halo effect – when a strength masks other weaknesses
- Overconfidence bias – when confidence exceeds reality
- Framing effects – when presentation shapes the decision
1. Authority bias
Authority bias occurs when the opinion of a senior leader or other prominent person carries more weight than is warranted by the available evidence.
Thus, once a senior figure has signaled support for an option early in a discussion, the conversation tends to coalesce around that position. Other perspectives are still voiced, but they are much less likely to shape the final outcome.
The pull of authority bias is deep-rooted in most people’s psyche: Stanley Milgram’s (1963) classic obedience experiments showed how readily people defer to authority, even against their own better judgment.
Example
A health system is prioritizing investments across competing service lines, and a senior clinical leader forcefully argues for expanding surgical capacity. Although the broader data points to preventive care as the higher-impact investment, discussion gravitates toward the surgical option and other alternatives receive less attention.
Solution
Gather participants’ inputs independently before group discussion begins. When participants record their views without seeing anyone else’s, the full range of expertise is captured rather than filtered through the first strong signal. Approaches like anonymous voting and structured preference elicitation are designed to support this kind of process.
2. Groupthink
Groupthink takes hold when the maintenance of alignment within a group becomes more important than the testing of alternative viewpoints and assumptions. Relevant concerns go unspoken because no one wants to be the person who slows things down, and the absence of objection is read as consensus.
Groupthink bias can also cause groups to take more extreme risks than necessary. First described by psychologist Irving Janis (1972) in his analysis of US foreign-policy fiascoes, groupthink is recognized as one of the most dangerous dynamics in organizational decision-making.
Example
A cross-functional committee is reviewing a shortlist of major projects, and one option has already gathered informal support. During the formal review, several members privately doubt its long-term value, and yet no one raises their concerns. The recommendation moves forward unchallenged – and only later, when results disappoint, do those doubts resurface.
Solution
Build structured challenge into the process so dissent is expected rather than exceptional. Collecting independent assessments in advance, or requiring risks and failure scenarios to be documented before a decision is finalized, surfaces concerns while they can still change the outcome. Iterative, anonymized methods such as the Delphi technique – developed at the RAND Corporation by Dalkey and Helmer (1963) to separate individual judgment from group pressure – are well suited to combatting this bias.
3. Sunk cost fallacy
The sunk cost fallacy leads organizations to keep investing in something because of what has already been spent, rather than what makes sense going forward. Past investment carries financial, political and emotional weight, which makes walking away difficult even when the original rationale no longer holds.
Arkes and Blumer’s (1985) foundational experiments demonstrated how strongly prior spending can distort subsequent choices.
Example
A large-scale implementation project is halfway through delivery and significantly over budget. Early results suggest the expected benefits are unlikely to materialize at the scale originally projected. However, leadership decides to continue, reasoning that stopping now would waste the time and money already invested – even though completing the project will cost more than the remaining benefits would justify.
Solution
Evaluate all initiatives – new and existing – against the same criteria on a regular basis, including at key milestones. If an existing project would not be funded under today’s conditions, that will become clear in the analysis, shifting the focus from money already spent to value still to be realized.
4. Anchoring bias
Anchoring bias occurs when early information – in the form of a number, an estimate or a first opinion – sets a reference point that colors everything that comes after it. Even when the anchor is arbitrary and completely unrelated to the decision context, later judgments are influenced by it.
Tversky and Kahneman (1974) identified anchoring as one of the core mechanisms people use to simplify complex judgments under uncertainty.
Example
A leadership team opens its review by discussing one project in detail, including an early estimate of expected impact. As further projects come up, that initial information becomes an informal benchmark, and each option is judged relative to it rather than on its own merits.
Solution
Structure how inputs are gathered so that no single data point dominates. Independent assessments before discussion prevent the first option evaluated from anchoring the others under consideration. Evaluating options through systematic pairwise comparisons is also helpful, because each option is compared directly against the others instead of against an arbitrary starting point.
5. Availability heuristic
The availability heuristic leads people to ascribe too much importance to information that is recent, vivid or easy to recall – a tendency Tversky and Kahneman (1973) documented in the early 1970s. In organizations, this bias means priorities can swing toward whatever issue is currently most visible, even when longer-term data points in a different direction.
Example
After a high-profile incident, leadership redirects resources to the specific issue it highlighted. The response is understandable, but broader analysis shows that other areas would deliver greater overall impact.
Solution
Define decision criteria in advance and apply them consistently. A stable framework lets new information be weighed on its merits without letting it override longer-term priorities – keeping the organization responsive without becoming reactive.
6. Confirmation bias
Confirmation bias shapes how evidence is gathered and interpreted once a preferred direction begins to form. Supporting data gets explored in detail, while risks and contradictory signals are acknowledged but examined less closely.
Confirmation bias is one of the most pervasive and well-documented tendencies in human reasoning – Raymond Nickerson’s (1998) review catalogues the many guises it takes.
Example
A team builds a business case for a digital program, emphasizing efficiency gains and strategic fit. Implementation risks and lessons from comparable past initiatives get lighter treatment. As the proposal advances, it becomes easier to justify than to question.
Solution
Require balanced evaluation as part of the decision-making process. For each claimed benefit, teams should identify the corresponding risks and the conditions under which the expected outcome might not materialize. A structured, multi-criteria framework helps, because it requires every criterion – favorable or not – to be addressed explicitly.
7. Choice overload
When organizations evaluate a large number of options across several dimensions, consistent judgment becomes increasingly difficult. As the volume of comparisons grows, attention becomes uneven. Some options get thorough review while others are assessed more quickly, and familiar choices tend to stand out simply because they are easier to process.
Example
A funding body reviews a large pool of proposals across multiple criteria. As the process wears on, scoring drifts and becomes inconsistent; previously funded organizations gain an implicit edge because evaluators already know them.
Solution
Break the decision into smaller, more manageable comparisons. Weighing a limited number of options at a time improves consistency and reduces cognitive strain. 1000minds’ PAPRIKA method (Hansen and Ombler, 2008) takes this approach to its logical conclusion, converting a complex multi-criteria decision into a series of simple pairwise trade-offs and combining them into a consistent overall ranking.
8. The halo effect
The halo effect occurs when a strong impression in one area colors judgment across the board. This bias was first documented by Edward Thorndike (1920), who found that a rating on one trait spilled over into unrelated ones. A project that excels on a visible or valued dimension gets treated as strong overall, even when other aspects have not been closely examined.
Example
A proposal aligns closely with a high-profile strategic priority and attracts strong early support. Because of that alignment, its financial viability and implementation plan receive less scrutiny than they should.
Solution
Assess each criterion independently, against clear definitions. Separating the evaluation dimension-by-dimension keeps strength in one area from obscuring weakness in another. Where possible, assign different subject-matter experts to different criteria according to their area of expertise.
9. Overconfidence bias
Overconfidence bias appears when decision-makers place more trust in their own judgment than the evidence supports. Experience is valuable, but confidence and accuracy do not always move together, particularly in complex or uncertain situations. In Vincent Berthet’s (2022) review of professional decision-making across management, finance, medicine and law, overconfidence was the single most recurrent bias.
Example
A panel of experts is asked to rank strategic initiatives. Each offers confident assessments, yet the rankings vary widely across the group. Even so, individuals tend to treat their own as the most reliable and attribute the differences to others lacking context – rather than as a prompt for closer examination.
Solution
Use methods that surface disagreement and require explicit trade-offs. When participants compare options systematically, inconsistencies become visible and can be discussed productively.
As popularized by Kahneman, Sibony and Sunstein (2021), a noise audit – collecting and comparing individual judgments before any discussion – is a powerful way to reveal how much expert assessments diverge and to focus the conversation on why.
10. Framing effects
Framing effects influence decisions based on how options are presented rather than the fundamental information they contain. The same initiative can attract enthusiasm or caution depending on whether it is described positively or negatively in terms of gains or losses and opportunities or risks.
Tversky and Kahneman (1981) demonstrated this bias with their well-known “Asian disease” problem, in which logically identical options drew opposite preferences depending only on whether outcomes were framed as lives saved or lives lost.
Example
A pilot program is presented to leadership with two metrics. Framed one way, it has "a 92% customer satisfaction rate." Framed another way, it "failed to meet expectations for 8% of customers." The data is identical, but the first framing builds confidence while the second raises concern – and whichever is presented first tends to set the tone for the discussion.
Solution
Where possible, present options through more than one framing – for example, both what is gained and what is given up – so that no single narrative dominates the discussion. Consistent decision criteria also help here: when decisions are anchored in defined criteria rather than narrative framing, the influence of presentation fades. Making trade-offs explicit ensures decisions reflect the substance of the options, not the way they happened to be described.
How cognitive biases affect decision-making in management
Cognitive biases in decision-making do not operate individually in isolation from each other. In management and leadership settings they interact with organizational culture, time pressure and political dynamics.
For example, a meeting can be shaped by groupthink, anchored to an early estimate and tilted by authority bias all at once.
Awareness of these biases alone is rarely enough to reduce their impacts. Training can help people recognize them, but biases tend to persist unless the decision-making process itself is designed to limit their impacts.
Structured approaches – e.g. independent inputs, application of explicit criteria and systematic comparisons – address the conditions that allows biases to take hold, rather than relying on individuals to catch them in the moment.
How structured decision-making reduces bias
Most of the process-based solutions to decision-making biases discussed in this article share these four basic principles:
- gather inputs independently
- define criteria clearly
- compare options systematically
- keep the reasoning transparent
The solutions discussed in this article are potentially available to all organizations. Though it's possible to perform them “by hand” (e.g. supported by spreadsheets), decision-support software such as 1000minds makes implementation – especially at scale – easier and more reliable.
1000minds is widely used across healthcare, government and the private sector. For example, the software was used by the World Health Organization to prioritize antibiotic-resistant pathogens for R&D – a very important decision characterized by limited resources, many competing criteria and genuine disagreements between global experts to be reconciled.
Multi-criteria decision analysis (MCDA)
1000minds supports MCDA, which, in a nutshell, is an analytical framework for evaluating options across multiple criteria, weighted according to their relative importance. By requiring each criterion to be assessed explicitly and independently, 1000minds’ MCDA reduces confirmation bias, the halo effect and framing effects.
Noise audits
Before a group makes decisions together, a 1000minds “noise audit” reveals how much individual judgments vary across groups – which often comes as a shock to expert participants. Identifying this judgmental variability is a productive starting point for structured discussion and consensus-building.
Structured trade-offs through pairwise comparisons
1000minds’ award-winning PAPRIKA method breaks down complex decisions into simple pairwise comparisons involving trade-offs between two criteria at a time, resulting in a valid and reliable ranking of the alternatives being evaluated. By limiting attention to just two things at a time, this approach directly counters choice overload, anchoring and the halo effect.
Independent inputs and anonymous voting
1000minds’ anonymous surveys and voting let participants contribute their own views before being exposed to other people’s. This participatory approach is one of the most effective ways to blunt authority bias and groupthink, because it captures genuine preferences before influence dynamics take over.
Conclusion
Decision-making biases are not a sign of poor leadership. They are intrinsic to how people process information, especially when decisions involve uncertainty, competing priorities and time pressure.
The organizations that handle biases best focus less on training individuals to “think better” and more on designing processes that naturally reduce the influence of biases. Independent input, clearly defined criteria, structured comparison and transparent reasoning all contribute to more consistent, defensible decisions.
The goal is not perfect objectivity. It is building a process in which the best available thinking has the best chance of shaping the outcome.
Frequently asked questions
What are decision-making biases?
Decision-making biases are systematic cognitive errors in how people process information and evaluate choices. They are predictable patterns – not random mistakes – that affect judgment, especially under complexity and uncertainty. In organizations, they influence how options are framed, which evidence is prioritized and how trade-offs are made.
What are the 10 most common decision-making biases?
The most common biases in organizational settings include: authority bias, groupthink, the sunk cost fallacy, anchoring bias, the availability heuristic, confirmation bias, choice overload, the halo effect, overconfidence bias and framing effects. Each bias affects a different stage of the decision process, from how information is gathered to how the final choice is justified.
How do cognitive biases affect decision-making in the workplace?
They distort how information is interpreted, which options receive attention and how risks are weighed. In practice, that can mean senior opinions carry disproportionate weight, past investments block rational reallocation or recent events override long-term data. These effects are amplified in groups, where individual biases interact with team dynamics.
Can decision-making biases be eliminated?
No – they are rooted in how human cognition works, so they cannot be fully removed. Their influence on organizational decisions can, however, be reduced substantially through structured processes such as independent input collection, clearly defined criteria and systematic comparison methods like multi-criteria decision analysis (MCDA).
References
H Arkes & C Blumer (1985), “The psychology of sunk cost”, Organizational Behavior and Human Decision Processes 35, 124-40 – doi.org/10.1016/0749-5978(85)90049-4
V Berthet (2022), “The impact of cognitive biases on professionals’ decision-making: a review of four occupational areas”, Frontiers in Psychology 12, 802439 – doi.org/10.3389/fpsyg.2021.802439
N Dalkey & O Helmer (1963), “An experimental application of the Delphi method to the use of experts”, Management Science 9, 458-67 – doi.org/10.1287/mnsc.9.3.458
P Hansen & F Ombler (2008), “A new method for scoring additive multi-attribute value models using pairwise rankings of alternatives”, Journal of Multi-Criteria Decision Analysis 15, 87-107 – doi.org/10.1002/mcda.428
I Janis (1972), Victims of Groupthink: A Psychological Study of Foreign-Policy Decisions and Fiascoes, Houghton Mifflin – Open Library
D Kahneman, O Sibony & C Sunstein (2021), Noise: A Flaw in Human Judgment, Little, Brown Spark – Open Library
S Milgram (1963), “Behavioral study of obedience”, Journal of Abnormal and Social Psychology 67, 371-8 – doi.org/10.1037/h0040525
R Nickerson (1998), “Confirmation bias: a ubiquitous phenomenon in many guises”, Review of General Psychology 2, 175-220 – doi.org/10.1037/1089-2680.2.2.175
E Thorndike (1920), “A constant error in psychological ratings”, Journal of Applied Psychology 4, 25-9 – doi.org/10.1037/h0071663
A Tversky & D Kahneman (1973), “Availability: a heuristic for judging frequency and probability”, Cognitive Psychology 5, 207-32 – doi.org/10.1016/0010-0285(73)90033-9
A Tversky & D Kahneman (1974), “Judgment under uncertainty: heuristics and biases”, Science 185, 1124-31 – doi.org/10.1126/science.185.4157.1124
A Tversky & D Kahneman (1981), “The framing of decisions and the psychology of choice”, Science 211, 453-8 – doi.org/10.1126/science.7455683
Share this post on:
