Skip to content

Decision-grade is the bar that separates research that ends a debate from research that adds to it.

The six criteria of decision-grade research

Decision-grade research meets six criteria. Each is necessary; none is sufficient on its own.

1. Framed around one decision

The research has a single, named decision at its centre. Everything in the engagement, sample, method, analysis, output, serves that decision. Research framed around a "theme" or a "category" can be useful, but it is not decision-grade. The discipline is to know which one of the two you're commissioning.

2. Built on primary evidence

Decision-grade research is grounded in primary sources, interviews with the actual buyers, operators, customers or experts who hold the truth in the category. Secondary sources are sanity checks; they are never the base case. Synthetic audiences are accelerants; they are not the evidence.

3. Triangulated across sources

Every meaningful finding rests on more than one source. Buyer interviews are cross-checked against operator interviews; both are cross-checked against existing data; data is cross-checked against public signals. Where sources disagree, the disagreement is named explicitly, and is often where the insight lives.

4. Surfaces dissent

Decision-grade research names the buyers, customers or experts who disagreed with the dominant pattern. Their characterisation matters, who they were, why they disagreed, what their disagreement implies. Dissenters are typically the early indicator of where the dominant pattern will break.

5. States confidence level explicitly

Every finding is annotated with a confidence level, high, medium or low, based on sample size, source quality and degree of triangulation. The verdict states overall confidence on the decision. Implicit confidence is not enough; the decision-maker needs to know how heavy each finding really is.

6. Defensible under scrutiny

The output has to survive structured scrutiny, by a board, an investment committee, an acquirer's diligence room, or an auditor's question. That means: the sample is named, the methodology is documented, the source trail is auditable, and the assumptions are explicit. Anything that cannot be defended is excluded.

Why the bar matters

Research that meets all six criteria is more expensive per page than research that doesn't. It is also more useful per decision. The arithmetic favours the bar when the decision is large.

The mistake most buyers make is to commission generic research at a discount and find, too late, that it cannot settle the decision they actually had. The savings disappear when the decision has to be re-researched, or worse, when a bad decision gets made on weak evidence.

At a glance

The six criteria.

  1. Framed around one decision.

    Sample, method, analysis and output all serve a single named decision.

  2. Built on primary evidence.

    Real buyers, operators and experts. Secondary sources are sanity checks; not the base case.

  3. Triangulated across sources.

    Findings rest on more than one source; disagreements are named.

  4. Surfaces dissent.

    Dissenting voices are named, characterised, and their disagreement read for implication.

  5. States confidence level.

    Every finding annotated. Verdict states overall confidence on the decision.

  6. Defensible under scrutiny.

    Sample, methodology, source trail and assumptions all auditable. Built for the diligence room.

Research that meets the bar costs more per page and less per decision.
Why decision-grade is the economically correct choice when the decision is large.
Frequently asked

Decision-grade research, questions.

What does 'decision-grade research' actually mean?

Research designed to settle one decision, built to the bar of standing up in front of a board, an investment committee or an acquirer's diligence room. The structure is inverted from generic research: it starts from the decision, not the brief.

How is decision-grade research different from board-ready research?

All decision-grade research should be board-ready. Not all board-ready research is decision-grade, some 'board-ready' research is well-presented but does not actually settle a decision. Decision-grade is the stricter bar.

What's the bar?

Six criteria: framed around one decision; built on primary evidence; triangulated across multiple sources; surfaces dissent; states confidence level explicitly; defensible under structured scrutiny. If any one is missing, the research isn't decision-grade.

Can synthetic AI audiences produce decision-grade output?

Not yet, and possibly not soon. Synthetic audiences are useful for early-stage hypothesis testing and message screening. They are not yet trustworthy as the sole evidence base for decisions where the buyer's actual reasoning matters.

Is decision-grade more expensive than generic research?

Per page of output, yes. Per decision settled, almost always cheaper, because the decision is what you're actually paying for, and generic research often doesn't settle it.

Book a call → Free 30 minutes. No obligation. We reply within one working day.
Get in touch

Tell us the decision you're trying to get right.

We reply within one working day.