In the boardrooms of elite organisations, few cognitive
patterns prove more seductive or more costly than desirability bias. Also known
as wishful thinking bias or outcome desirability bias, it is the systematic
tendency for individuals to overestimate the probability of outcomes they find
desirable and to underestimate the probability of outcomes they find
undesirable. Desire itself becomes a causal input into probability judgement.
Preferences shape expectations.
This is not ordinary optimism or overconfidence. It is a
specific form of motivated reasoning in which the attractiveness of a future
state inflates its judged likelihood. Research in judgement and decision making
has repeatedly demonstrated the effect across laboratory and field settings. In
controlled studies, desirability can shift probability estimates by 10 to 20
percentage points or more. In multi-round Delphi foresight exercises involving
domain experts, participants consistently assign higher probabilities to
projections they rate as more desirable, even after iterative feedback. One
widely cited review of the literature shows that the bias is robust when
outcomes are personally or organisationally preferred, and that it operates
through multiple mechanisms including selective information search, biased
interpretation of ambiguous evidence, and the drive for cognitive consistency
between preference and prediction.
The dangers are clear and measurable. Distorted forecasts
lead to misallocated capital, inadequate contingency planning, inflated
valuations, delayed risk mitigation and strategic overcommitment. In project
environments, related optimism effects contribute to average cost overruns of
20 to 50 percent and schedule slips of similar magnitude. In investment and
M&A contexts, the correlation between rated desirability of a deal thesis
and the probability assigned to successful integration or returns is positive
and substantial. Organisations that fail to interrupt the bias systematically
under-prepare for downside scenarios while over-investing in preferred
narratives. The result is value destruction that only becomes fully visible
after the fact.
Desirability bias surfaces across industries wherever
high-stakes forecasting, long-horizon strategy and emotional or reputational
stakes coincide. Technology platforms, private equity venture capital, real
estate and proptech, aerospace and defence, pharmaceuticals and biotech, energy
transition projects, financial services, consumer brands and professional
sports all exhibit it. It is especially acute among elite individuals and industry/company
leaders, because status, legacy, narrative coherence and competitive identity
amplify the desirability of certain futures.
Three Illustrative Cases and Their Parallels
To illustrate the bias in action we examine three
organisations. Two are widely known and drawn from different industries. The
third is anonymised. It is the engagement in which our firm worked directly
with decision makers to diagnose and curtail the bias.
Case 1: Meta
Platforms and the Metaverse Bet (Technology)
In 2021 Facebook rebranded as Meta Platforms and committed
itself publicly and financially to the metaverse as the successor to the mobile
internet. Leadership articulated a highly desirable vision: hundreds of
millions or billions of users, hundreds of billions of dollars in digital
commerce, and a foundational shift in how people work, socialise and create.
Wall Street and internal forecasts attached multi-trillion-dollar total
addressable market estimates to the concept. Reality Labs, the division charged
with delivery, subsequently recorded operating losses in the range of 70 to 90
billion dollars over the following years. User adoption of flagship virtual
environments remained far below projections. In 2025 and 2026 the company began
significant reallocation of resources and workforce reductions within the
metaverse-related activities while pivoting capital toward artificial
intelligence infrastructure.
The desirability of becoming the company that defined the
next computing platform was intense. That desirability systematically inflated
probability estimates of mass adoption, technological readiness and commercial
returns, while dampening attention to early signals of weak engagement and high
friction. The parallel is classic: a preferred future state caused probability
judgements to diverge from base-rate evidence and early performance data.
Case 2: WeWork and
the Shared Workspace Vision (Proptech / Commercial Real Estate)
WeWork rose from a flexible office provider to a
private-market valuation of approximately 47 billion dollars by early 2019,
with some bank pitches exploring figures above 100 billion. The narrative was
highly desirable: a technology-enabled “physical social network,”
community-driven culture, and a spiritual redefinition of the workplace that
would disrupt traditional commercial real estate. SoftBank and other
sophisticated investors poured substantial capital into the story. Fundamentals
(long-term lease obligations versus short-term membership revenue, persistent
losses, and unit economics) received less weight relative to the preferred
vision. When the company attempted a public listing the valuation collapsed.
SoftBank later took control at a valuation in the single-digit billions,
representing a write-down of more than 80 percent from the peak private mark.
Here the desirability of the glamorous, identity-affirming
growth story and the associated status of backing a category-defining company
again elevated judged probabilities of success and scalability. Objective risk
factors (capital intensity, lease duration mismatch, governance issues) were
underweighted. The parallel with the Meta case is structural: in both instances
a compelling preferred future systematically distorted probability assessment
and capital allocation among highly sophisticated decision makers.
Case 3: Anonymised
Client Engagement (Industrial Technology and Advanced Manufacturing)
The third case is one in which our firm was retained. We
refer to the organisation as “Client Gamma,” a prominent European industrial
technology group with global operations, substantial R&D intensity and
ambitions in digital and sustainability-related markets. Leadership and the
board had developed a strong preference for a multi-year strategic platform
investment: a combination of organic capability build and a large complementary
acquisition that would position the company as a leader in an emerging high-visibility
segment. The thesis was attractive on multiple levels (technological prestige,
ESG narrative alignment, competitive signalling, and potential long-term
growth). Internal probability estimates for successful execution, synergy
realisation and market timing clustered in the 70 to 85 percent range.
Independent base-rate analysis of comparable industrial transformations and
acquisitions suggested a more realistic range closer to 35 to 50 percent under
similar conditions.
Desirability bias was operating at board and executive
level. Preferred outcomes were receiving inflated likelihood weightings.
Alternative scenarios and contingency costs were under-specified. The
engagement allowed us to work directly with the decision makers to interrupt
the pattern before irreversible capital commitments were finalised.
How We
Work with Elite Decision Makers to Interrupt Desirability Bias
Our methodology is diagnostic, collaborative and
measurement-oriented. It is designed for boards, CEOs, CFOs, strategy teams and
investment committees.
1. Confidential
Diagnostic and Bias Mapping
We begin with
structured interviews, decision process mapping and review of recent investment
cases, forecasts and strategic papers. We quantify the gap between stated
probabilities and reference-class forecasts. Desirability ratings of key
outcomes are elicited separately from probability estimates so the correlation
can be measured directly. In the Client Gamma engagement this revealed a
statistically significant positive relationship between rated desirability and
assigned probability, independent of objective evidence strength.
2. Probability
Elicitation and Pre-Mortem Protocols
We replace informal
“we believe it is highly likely” language with structured techniques: reference
class forecasting, premortem exercises, independent red-team probability
assessments, etc. Decision makers are asked to estimate probabilities under
both desirable and undesirable framings to surface the bias explicitly.
3. Process Redesign
and Decision Hygiene Rules
We install
pre-commitment frameworks: mandatory alternative-option generation, required
downside scenarios with explicit probability weightings, and separation of
preference statements from likelihood statements in board papers. Information
search protocols are adjusted to counteract selective exposure to confirmatory
or desirability-aligned data.
4. Board and
Executive Capability Building
Interactive
workshops and decision journals train participants to recognise the bias in
real time. We use calibrated scoring (Brier scores or similar) on historical
predictions to create personal feedback loops. In Client Gamma, board members
and the executive committee completed a multi-session programme that measurably
improved calibration.
5. Implementation
Support and Measurement
We embed revised
templates, dashboards and stage-gate criteria. Leading indicators
(probability-desirability correlation, quality of scenario coverage, decision
cycle time) and lagging indicators (capital efficiency, forecast accuracy,
contingency usage) are tracked. Quarterly reviews with the board ensure
accountability.
In the Client Gamma engagement the process produced clear
shifts. Probability estimates for the preferred platform were revised downward
by an average of 22 percentage points after structured interventions. The board
ultimately approved a more modular, staged investment approach with explicit
kill criteria rather than a single large commitment. Subsequent performance
tracking showed improved capital discipline and earlier course correction when
market signals diverged from the original thesis. Decision quality scores
(process adherence and calibration) rose from a baseline of 58 percent to 86
percent within twelve months.
Empirical Patterns
and Broader Correlations
Across published studies and our own anonymised portfolio
data, several regularities appear:
- The higher the rated desirability of an outcome, the
larger the average upward distortion in probability estimates (often 12 to 25
percentage points in organisational settings).
- Expert groups are not immune. Delphi-style foresight work
shows residual desirability effects even after multiple rounds.
- Organisations that introduce structured bias interruption
protocols typically record 15 to 30 percent improvements in forecast accuracy
and 20 to 40 percent reductions in large-project overruns within two to three
years.
- Boards that treat desirability bias as a governable risk
rather than an inevitable human trait demonstrate higher long-term capital
efficiency and fewer “strategic surprises.”
The Meta and WeWork examples illustrate the extreme end of
the spectrum when the bias operates unchecked at scale. The Client Gamma case
demonstrates that the same underlying mechanism can be interrupted with
disciplined process before the costs compound.
Conclusion:
From Aspiration to Calibrated Judgement
Desirability bias is not a character flaw. It is a
predictable feature of human cognition that becomes especially powerful when
elite individuals and boards confront futures that are status-enhancing,
identity-affirming or competitively seductive. Left unaddressed, it converts
preferred stories into overweighted probabilities and converts capital into
regret.
Organisations that treat bias interruption as a core
governance and strategy capability convert the same human tendency into a
manageable variable. They preserve the energy of ambition while anchoring
decisions in evidence, base rates and explicit probability discipline. The
result is not diminished vision. It is vision that survives contact with
reality.
If your board or leadership team is evaluating major
strategic investments, platform bets or multi-year transformations, a
structured diagnostic of desirability bias can reveal hidden risk and create
immediate decision advantage. Our firm specialises in exactly this work with
elite individuals and governing bodies. We would welcome a confidential
conversation about how the approach can be tailored to your context.
*This article is based on established research in judgement
and decision making, publicly observed corporate outcomes, and anonymised
client experience. Client identities and proprietary details remain
confidential.*