The Translation Problem
Why strong research fails to influence decisions
Strong research often fails to change decisions — not because it is wrong, but because it stays in the form in which it was produced. A paper is written for reviewers and the scholarly record. A decision brief is written for someone who needs to act.
— Rankine Innovation Lab · Knowledge Hub
A decision brief solves a translation problem. It reorganises the most decision-relevant parts of a research paper for readers who need clarity more than academic form. This does not mean oversimplifying the work — it means translating it responsibly. The original researchers should still recognise what their work says. But the decision-maker should understand what it means for them.
The distinction between a brief and a summary is critical and consistently missed. A summary reproduces content in a shorter form. A decision brief reconstructs it around a different question: not "what did this paper say?" but "what should someone do differently as a result of this evidence?"
Conceptual Foundation
A decision brief is not a shorter summary
Understanding what distinguishes a brief from a summary is the first prerequisite for writing a good one. The differences are structural, not stylistic — they reflect a genuinely different purpose.
Written for the person who must act
Begins with the decision context, not the paper's abstract
Extracts only the evidence relevant to the specific decision
Names implications, options, and recommended next moves
States clearly what the evidence does not yet support
Can be acted on without reading the original paper
Written around the paper's own structure
Follows the paper's sequence — introduction through conclusion
Reproduces findings in shorter form without prioritising them
Stops at "what was found" without reaching "so what"
Leaves the decision-maker to draw their own implications
May be accurate without being decision-ready
Implementation Sequence
The six-step translation process
Each step in this sequence has a distinct purpose. Skipping steps does not accelerate the process — it shifts the failure mode from "brief not started" to "brief that looks finished but doesn't work." The order matters because purpose must precede extraction, which must precede framing.
Define the decision context before opening the paper
Start by writing down the decision, the audience, and the practical question the brief needs to support. This takes ten minutes and changes every subsequent choice. Without it, the extraction process is directionless and the brief will serve nobody in particular.
"What must this person decide, and what do they need to trust that decision?"
Extract the decision-relevant core
From the paper, identify and pull out: the problem statement, the main finding, the evidence boundary (what conditions the finding applies to), and the most direct practical implication. Everything else can be referenced but not reproduced.
Rebuild structure around the decision, not the paper
Use the six-section brief structure: context, what the research examined, what it found, what this means, limits and cautions, and options or next steps. Do not follow the paper's own section order — that structure was designed for peer reviewers, not practitioners.
Translate without flattening
Remove avoidable complexity while preserving caveats and conditions. The test is whether a researcher would still recognise their finding — if the translation has stripped away the conditions under which it holds, it has become a distortion, not a simplification.
"Uncertainty retained is a sign of quality, not weakness."
Add a decision layer — options, not conclusions
Turn findings into options, pilot logic, no-go guidance, or monitoring requirements. A brief that names implications but not actions leaves the decision-maker one step short of usefulness. Frame options as contingent on conditions, not as fixed recommendations.
Review against both tests before sharing
Apply the dual test: would the original researcher recognise their work in this brief? And would the decision-maker be able to act on it without first reading the paper? Both must be true. A brief that passes only one test is either accurate-but-unusable or usable-but-distorted.
Practical Tool
The six-section brief structure that works
This section structure is not a template to fill in — it is a sequencing logic. Each section does a specific job that the preceding section makes possible. The order reflects how decision-makers actually absorb evidence under time pressure.
1
Decision Context
One paragraph explaining the practical question and audience. This is what justifies the brief's existence and tells the reader why they are holding it. Without this, the reader has no anchor for interpreting what follows.
Why this, why now, for whom
2
What the Research Examined
A short, precise description of the study — its scope, method, and conditions. Not a literature review. Just enough to let the reader assess whether the evidence is relevant to their context.
Scope and method
3
What It Found
The most decision-relevant result in plain language. Not the full findings section — only the finding that connects most directly to the decision context defined in section one. Other findings can be referenced but not foregrounded.
The headline evidence
4
What This Means
Practical implications for the target audience — what the evidence suggests about how to act, what to change, or what to investigate further. This is where translation becomes interpretation, and where the author's domain knowledge matters most.
Implications and significance
5
Limits and Cautions
What the evidence does not yet justify. This section is where a brief earns its credibility. Stating limits clearly is not defensive — it is what allows the decision-maker to trust the rest of the brief.
What should not be claimed
6
Options or Next Steps
Possible actions framed clearly and proportionately. Frame as options with conditions attached, not as recommendations. Name who could act on each option, and what additional evidence would be needed to increase confidence.
Conditional actions
Extraction Framework
Five questions to answer from the paper first
Before writing a single word of the brief, work through these five questions using only the paper as a source. If any cannot be answered clearly, the brief is not yet ready to be written — or the paper may not be as decision-relevant as assumed.
01
What practical issue does this research speak to?
If the answer is vague ("AI in general" or "sustainability"), narrow it further until it names a specific decision or workflow problem.
02
What does the paper actually show — not claim?
Distinguish between the finding and the interpretation the authors place on it. Both belong in the brief, but they should not be conflated.
03
Under what conditions does this finding hold?
Every evidence claim has boundary conditions — the specific context, scale, population, or setting in which it was observed. Name them precisely; they constrain what can be claimed.
04
Why does this finding matter to the specific decision audience?
This is the translation step. What changes for the decision-maker if this finding is accurate? What does it open, close, or complicate in their situation?
05
What actions, pilots, or further questions does this evidence open — and what does it not yet support?
Both halves of this question matter equally. A brief that only names what the evidence supports invites overclaiming. The limit statement is what makes the options credible.
Critical Awareness
Three mistakes that undermine the brief's value
These three failure modes account for most poor decision briefs. Each one produces a different kind of unusability — and each is predictable enough to design against in advance.
1
📋
Confusing a brief with a summary
A summary reproduces content; a brief supports a decision. If the document follows the paper's structure, it has become a summary — regardless of how short or clearly written it is. Structure reveals purpose.
2
🎭
Stripping uncertainty to appear decisive
A brief that removes all caveats to sound confident has become a distortion. Uncertainty preserved is what allows the decision-maker to calibrate their confidence appropriately. A brief without limits is not trusted — it is suspected.
3
📏
Making the brief too long
If the brief recreates the density of the paper, it loses its value. The brief should be lean enough to support action and strong enough to protect rigour. If a decision-maker needs to read it twice slowly, it has not been translated — it has been reformatted.
Quality Gate
Before sharing — apply this completion check
Before circulating a decision brief, work through these five checks. If any answer is weak, revise the brief before sharing it. A weak brief shared widely is harder to correct than one caught before distribution.
?
Is the decision context stated clearly — and does everything else in the brief serve that context?
?
Would the original researchers recognise their finding in the way it has been translated here?
?
Are the limits and cautions as prominent as the implications — or have they been buried at the end?
?
Are the options proportionate to the evidence — or do they claim more than the research supports?
?
Can the target reader act on this brief without first reading the original paper?
References & Source Base
- Rankine Innovation Lab Knowledge Hub research brief: Playbook direction for research-to-decision translation, with editorial framing for the cross-disciplinary innovation domain.
- Cross-link: What Repeated Reading of GenAI Papers Is Revealing — Rankine Knowledge Hub. Provides an applied example of reading research for decision-relevance rather than academic form.
- Cross-link: AI Adoption Readiness for Research Teams — Rankine Knowledge Hub. Shows the kind of decision context that a brief built on AI evidence might be designed to support.
- Cross-link: RAG for STEM Decision Support — Rankine Knowledge Hub. Illustrates how retrieved evidence can be translated into usable decision-support language.