Citation: Csaszar, F. A. and Rhee, L. (2026). The power and limits of distributed representations in strategic decision-making. Strategy Science 11(2) 209–228. doi:10.1287/stsc.2023.0023
Cite
Paper highlights
Consider a leadership team deciding whether to launch a new product. Engineering understands technical feasibility, marketing understands demand, and finance understands cost, but no one has the full picture. The team must decide whether to trust one specialist, require everyone to agree, average their judgments, or rely on a generalist. This paper explains when each arrangement produces the most accurate decision.
Each specialist sees only part of the problem. When their judgments are combined, the organization creates a distributed representation: a collective understanding that no member possesses alone. Its quality depends on the fit among what each person knows, how their judgments are combined, and the decision environment. The same team can perform differently when good projects are rare, one source of information dominates, or members gain experience.
How the model works
The formal project-screening model compares single-cue specialists, a generalist who learns several cues, and two structures combining specialists: Averaging and Unanimity. It varies experience, uncertainty, environmental complexity, the prevalence of good projects, and whether one cue is much more informative than the others.
Main results
- A highly experienced generalist performs best across the modeled conditions, but acquiring both breadth and depth requires substantial experience.
- A specialist is best in simple environments dominated by the cue that specialist knows.
- Unanimity works well when good projects and experience are scarce because it screens conservatively.
- Averaging is the most robust default across most other conditions.
- Under uncertainty, matching experienced specialists through one generalist can require unrealistically large amounts of experience.
Why it matters
- An organization’s representation resides partly in what each member sees and partly in the rule used to combine their judgments.
- Unanimity is useful when good projects are rare and false positives are costly; averaging is more robust across less extreme conditions.
- A deeply experienced generalist can outperform the modeled teams, but distributed representation lets organizations combine specialist depth that no readily available individual possesses.
How to use this paper
Cite this for
- A formal theory of distributed representations in strategic decision-making.
- A comparison of specialists, generalists, averaging, and unanimity in project screening.
- Conditions under which distributed partial models help or hurt collective judgment.
Useful for teaching
- Why no one person may hold the whole strategic representation.
- How the rule for combining specialists becomes part of the organization’s representation.
- Why good-project scarcity, cue dominance, experience, and uncertainty change the best aggregation rule.
Careful claim
The model shows that distributed representations depend on the environment, experience, cue structure, and aggregation rule; averaging is robust in many modeled conditions, but not universally best.
Abstract
This paper develops a formal theory of distributed representations—collective cognitive models that emerge when organizations aggregate the simplified mental models of multiple members—and examines when they enhance or hinder strategic decision-making.
We extend Brunswik’s lens model to multiple decision-makers and introduce “decision boundaries” from machine learning to explain how aggregation structures interact with individual internal representations across varying task environments. Using a mathematical model of project screening, we compare two prototypical aggregation rules (Averaging and Unanimity) against individual Specialists (single-cue experts) and Generalists (multi-cue learners) across various environments and levels of experience.
Our analysis reveals that effectiveness depends critically on the three-way interaction between internal representations, aggregation structure, and environmental conditions: Specialists excel when one cue dominates; Unanimity guards against errors when good projects are rare and decision-makers lack experience; Averaging delivers robust performance across most settings; while only highly experienced Generalists outperform distributed representations, though such individuals are scarce in practice. These findings advance microfoundations by linking individual cognition and organizational aggregation, enrich the attention-based view by showing how cognitive processing and aggregation matter beyond attention allocation, and offer actionable guidance for designing decision processes under strategic uncertainty.
Other links
Last updated 2026-06-21