Why a Thorough Market Research Report Changed Nothing

Last updated: 31 August 2026

You paid for it. It was thorough, it was on time, the methodology was sound, and nothing happened. No clearer prioritisation, no stronger conviction, just interesting insights. The obvious conclusion is that you hired the wrong firm.

The evidence points somewhere less satisfying. Most of the reasons research fails to change a decision sit structurally on the buyer's side, and in a large share of cases a better supplier would have produced exactly the same non-event.

I should say plainly that this is against my interest. I sell research and go-to-market work. The argument below says that buying more of it, or buying better of it, is frequently not the fix.

Why did a thorough market research report change nothing?

That framing is the most useful thing I found, because it explains the specific shape of the disappointment. You did not buy a decision. You bought information, which is one requirement of six.

If the frame was wrong — if the question asked was not the question that mattered — better information about the wrong question is still the wrong question, answered well. A market size built to impress rather than to decide fails here for the same reason. If no real alternatives were on the table, information cannot generate them. And if nobody had committed to act on any particular finding, there was never a mechanism by which a finding could become an action.

What does "it changed nothing" actually mean?

Weiss gave the thing you experienced a name, and treated it as the norm.

Weiss also named symbolic use, where research justifies a position already held. That one matters later. The honest translation of "it changed nothing" is narrower than it feels: it rarely means the research was wrong, only that it did not reach the rarer category of use.

One caution I want to put early rather than bury. This literature is drawn almost entirely from healthcare, public policy and large organisations. It has essentially never been connected to a startup or mid-market company buying a commercial market report. Every transfer in this article is an inference, and I have tried to mark them as such rather than smuggle them past you.

Why did the research fail to drive action?

The Oliver review is the largest synthesis available on this question, and its findings are unkind to the way most research is bought. The review's own summary: the most frequently reported barriers were "poor access to good quality relevant research, and lack of timely research output", and the most frequently reported facilitators were "collaboration between researchers and policymakers, and improved relationships and skills". What does not predict use is how good the document is.

The pattern repeats elsewhere. Studies of university research use in government found that characteristics of the research product are weak predictors of use while the user's context is the strongest, and a study of research use in child protection surveying 447 administrators and practitioners found only 18% reported using research-based knowledge on a frequent basis, with relational capital between researchers and users and perceived usefulness the two factors most strongly related to utilisation.

Pfeffer and Sutton's The Knowing-Doing Gap (2000) supplies the mechanism that makes this feel so familiar: talk, plans, analysis, meetings and presentations become substitutes for action. They call it the smart talk trap. Commissioning research is one of the most respectable ways to participate in it, and it is why so much go-to-market activity feels busy and inert at once.

What are the common reasons stakeholders ignore research?

The uncomfortable one is motivated reasoning. Taber and Lodge's experimental work on motivated scepticism found subjects devalued arguments that conflicted with their prior preferences irrespective of argument quality, and spent measurably more effort counter-arguing disconfirming evidence than supporting evidence. The setting was political attitudes rather than boardrooms, so the transfer is an inference — but escalation-of-commitment and cognitive-dissonance research point the same way, and every operator has watched a methodology suddenly come under intense scrutiny at the exact moment its conclusions became inconvenient.

Pfeffer and Salancik documented the organisational version in 1974: information developed to justify a decision that was already desired. That is Weiss's symbolic use, seen from inside.

Notice what is absent from that list. Not one of the seven failure modes is "the research was inaccurate". Three sit squarely with the buyer, and the remaining four are shared. This is the whole reason the "better supplier" reflex misfires: you are replacing the one component that was not broken.

What makes a market research report actionable?

Here is where the evidence gets genuinely thin, and I would rather say so than dress it up.

The formatting advice in this industry is confident and largely unfounded. The familiar rules — an executive summary at 5 to 10% of length, one to two pages, 300 to 500 words — appear across writing guides and vendor blogs with no empirical link between length and decision uptake, and they do not agree with each other. Whether recommendations help or trigger rejection has no evidence in a commercial setting, and neither does visual presentation. These are conventions wearing the clothes of findings.

The one place decision-forcing format has actually been studied is clinical guidance, where the stakes forced the profession to formalise it. GRADE's Evidence-to-Decision framework, used across 15 international guideline panels, requires every recommendation to carry both a direction and a strength, alongside an explicit rating of how certain the underlying evidence is. As its authors put it, the framework provides "an explicit record of the judgments" that determine where a recommendation lands and how hard it pushes.

And even there, the discipline is uneven. A 2022 review in the Journal of Clinical Epidemiology examined 68 guidance documents: 93% used a structured framework to assess certainty of evidence and 88% used one to rate recommendation strength, but only 66% explicitly stated the process by which their recommendations were formed. If a third of clinical guidance does not say how it decided, the bar your research supplier is being held to is not a high one.

A user test of evidence summaries adds the complements: jargon and length hinder use, and readers wanted actionable recommendations the format did not supply.

Transplanting that into a commercial market report gives you a shape worth asking for: an explicit recommendation with a confidence level, each option with its expected consequences, the disconfirming evidence stated rather than buried, a named decision owner, a deadline, and the cost of doing nothing. I could find no decision-forcing format native to commercial market research. The transfer is an inference, and it is the best available.

How do I turn findings into decisions?

Everything above points at the brief rather than the report. This is the version I would now insist on before taking research work, and would want as a buyer:

1. The decision. One sentence: "we must decide whether to ___."
2. The decision owner. A named individual with the authority to act, not a committee.
3. The deadline. The date the decision must be made.
4. The options. The two to four courses of action genuinely under consideration.
5. The pre-commitment map. "If the research shows A we do option 1; if it shows B we do option 2." Written before fieldwork starts.
6. The evidence threshold. The confidence level sufficient to act, which is never certainty.
7. The cost of inaction. What happens if no decision gets made.
8. The disconfirming test. What finding would prove the currently preferred option wrong.

Point five is the load-bearing one, and the one almost nobody writes down. A pre-commitment converts the report from an input into a trigger.

A pre-mortem is the natural companion: imagine the decision has already failed, and work backwards through why. The technique rests on prospective hindsight, where imagining an event has already occurred improved people's identification of reasons for the outcome by roughly 30% in the underlying 1989 study. That figure concerns risk identification rather than research uptake, so treat it as motivation for the practice, not proof that it lifts research use.

Be honest about what is missing here: there is no controlled test showing that a decision-linked brief improves research uptake in a commercial setting. The reasoning is sound and the components are individually evidenced. The compound claim is not.

How do you know whether the research was used at all?

This should change how you buy. Research use is measured mostly by asking people whether they used research, which is exactly the kind of question people answer badly. The attribution problem is well documented: a decision can be influenced without anyone reporting it, and tracing an output to an outcome resists clean causal claims.

So "did the report work?" is close to unanswerable after the fact. "Did the rule we agreed in advance fire?" is answerable by anyone, in a minute, and it is the reason to write the rule down beforehand.

What is the biggest problem facing market research today?

Read the industry's own diagnosis with the incentive in view.

Every framing is plausible. Every one also locates the failure exactly where the source sells a remedy — activation software, research tooling, transformation consulting, data infrastructure. None locates it at commissioning, where the literature puts it, and where no product reaches.

Two widely quoted figures do not survive contact with their sources. A claim that 30 to 40% of enterprise reports add little or no value, attributed to McKinsey, traces to a single trade article with no study, title or date, and I could locate no McKinsey publication with that wording. A claim that up to a third of research data goes unused traces to an unnamed study by a survey firm. The traceable figure is the 15%, and even its commonly cited year is wrong: the fieldwork was 2024.

I could find no quantified rate of research non-use specific to commercial market research. Everything on offer is either policy and healthcare data, or vendor surveys with undisclosed samples behind gated reports. If someone quotes you a percentage of research that goes unused, ask which study, and watch what happens.

Frequently asked questions

Why did the market research fail to drive action?

Most likely because no decision was attached to it at commissioning. Oliver and colleagues' 2014 systematic review of 145 studies found timeliness and relationships, not report quality, drive research use, and Weiss showed in 1979 that most research produces conceptual influence rather than instrumental action. A polished, correct report with no decision owner behind it is inert by design rather than by defect.

How can I turn research findings into business decisions?

Define the decision, the owner, the deadline and the options before fieldwork, and pre-commit to which finding triggers which option. That maps to the commitment requirement in Decision Quality (2016). Run a pre-mortem to surface objections early, and measure success by whether the pre-agreed rule actually fired rather than by whether people say the report was useful.

What makes a market research report actionable?

An explicit recommendation with a stated confidence level, options with their expected consequences, and disconfirming evidence stated plainly — the structure of GRADE's Evidence-to-Decision framework, used across 15 international guideline panels. Even in clinical guidance the discipline is uneven: a 2022 review of 68 guidance documents found only 66% explicitly stated how their recommendations were formed. No equivalent format is native to commercial market research.

What are common reasons stakeholders ignore research?

Motivated reasoning, where people devalue evidence conflicting with prior beliefs regardless of its quality; research commissioned to ratify a decision already made; poor timing, where the answer lands after the decision window closes; and the knowing-doing gap, where producing analysis substitutes for taking action.

Was the research supplier at fault?

Sometimes, but less often than it feels. Of the seven common failure modes, scoping, ownership and confirmation sit with the buyer, while timing, format, conviction and actionability are shared. None of the seven is inaccuracy. A supplier is at fault for accepting a brief with no decision, no owner and no deadline attached, which is a different failure from doing the research badly.

How do I write a research brief that leads to a decision?

State the decision in one sentence, name the individual who will make it, fix the date, list the two to four options actually under consideration, and write the pre-commitment map before fieldwork: if the research shows A we do this, if B we do that. Add the evidence threshold sufficient to act, the cost of inaction, and the finding that would prove your preferred option wrong.

What to do with this

If a report just landed and changed nothing, the diagnostic is quick. Go back to the brief and look for four things: a decision stated as a decision, a named owner, a date, and a written rule connecting findings to options. If any is missing, you have your answer, and a second supplier would not have supplied it.

Then do the cheap version of the fix. Before the next piece of research — whether you buy it or build the numbers yourself — write the eight-line brief above and get the decision owner to sign the pre-commitment map. It costs an afternoon and it is the only intervention here that acts on the link that was actually open.

And hold the finding lightly in one specific way. Everything above is inferred from healthcare and public policy, because that is where this was studied. Nobody has measured it on a startup buying a market report. That gap is the honest state of the evidence, and it is worth more to you than a confident number would be.

---

Sources

Decision and utilisation literature

• Spetzler, Winter and Meyer, Decision Quality (Wiley, 2016) — the six requirements, from the Stanford decision-analysis tradition. A practitioner framework, not an effect-size study, published by authors with a consulting interest in it.
• Carol H. Weiss, "The Many Meanings of Research Utilization", Public Administration Review (1979) — instrumental, conceptual, symbolic and enlightenment use.
• Oliver, Innvær, Lorenc, Woodman and Thomas, systematic review of barriers and facilitators of evidence use, BMC Health Services Research (2014) — 145 studies.
• Comparison of determinants of research knowledge utilisation in child protection (2010) — survey, n=447 (83 administrators, 364 practitioners).
• Landry, Amara and Lamari on university-research utilisation in government; Wallin and Estabrooks' review of 41 measurement studies.
• Pfeffer and Sutton, The Knowing-Doing Gap (Harvard Business School Press, 2000); Rousseau, "Is There Such a Thing as Evidence-Based Management?", Academy of Management Review (2006).
• Pfeffer and Salancik on decision-making as a political process (1974); Taber and Lodge on motivated scepticism (American Journal of Political Science, 2006); Kunda on motivated reasoning (1990).
• Bornmann, measuring impact in research evaluations (2014) — the attribution problem.

Decision-forcing formats

• GRADE Evidence-to-Decision frameworks, used across 15 international guideline panels; Meneses-Echavez et al., "Evidence to decision frameworks enabled structured and explicit development of healthcare recommendations", Journal of Clinical Epidemiology (2022), analysing 68 guidance documents.
• User testing of evidence-summary formats (2018).
• Klein, "Performing a Project Premortem", Harvard Business Review (2007), drawing on Mitchell, Russo and Pennington's prospective-hindsight work (1989).

Industry sources, each primary to itself and commercially interested

• McKinsey Growth Leaders Mindset Survey (fielded June to July 2024, n=500) — the 15% figure, surfaced in McKinsey Quarterly, January 2025.
• Insights-platform, research-tooling, marketing-effectiveness and data-intelligence publications from 2026, each locating the failure in the remedy it sells. Sample sizes for two of them sit behind gated reports and are not publicly disclosed.

What I could not establish

• No quantified rate of research non-use exists for commercial market research; every available figure is drawn from policy and healthcare settings or from vendor surveys with undisclosed samples.
• The widely circulated "30 to 40% of reports add little or no value", attributed to McKinsey, traces to a single trade article with no study, title or date, and no matching McKinsey publication was locatable.
• No controlled test shows that a decision-linked brief or a pre-commitment improves research uptake in a commercial setting.
• Three foundational papers here — Weiss (1979), Taber and Lodge (2006), and Mitchell, Russo and Pennington (1989) — were confirmed through citing and secondary sources rather than opened at the publisher's PDF.