Insight Teams Can Influence Decisions Without Proving It

Only 17% of insight professionals in a new international survey said they consistently measure research ROI or business impact. Yet more than half said their work had been cited in senior leadership decisions during the previous quarter.

That gap should not be read as evidence that research lacks value. It suggests something more specific: insight teams are often present at consequential moments, but cannot always show the route from evidence to changed commercial action.

The finding comes from Voxpopme and the Market Research Institute International’s *State of Insights* study, based on 126 experienced practitioners across client-side and supplier organisations. It is a small, self-selecting sample rather than a census of the profession, but its contradiction will feel familiar to many teams. A study can shape a proposition, prevent a poor launch, reframe a target audience or expose a risky assumption — and still leave little formal trace once the presentation is over.

This is not solved by forcing every project into a financial-return calculation. Research is often one input among several, alongside operational constraints, competitive pressure, commercial judgment and available investment. Claiming that a survey alone generated a revenue outcome is usually implausible.

The more useful task is to make research’s contribution to a decision visible while it can still be seen.

Influence is not the same as impact

There is a tendency to treat senior attention as proof of impact. A well-attended debrief, a request for a follow-up or a reference in a leadership meeting are all encouraging signals. None, on their own, proves that the organisation learned something it acted upon.

Equally, a project does not have to produce a dramatic change of direction to matter. Confirming that a proposed route remains sound may avoid costly delay. Finding that a buyer segment is too heterogeneous for a single message can improve a campaign before media money is committed. In B2B markets, evidence that procurement teams and day-to-day users are weighing different criteria can prevent a sales proposition being designed for the wrong audience.

Impact therefore has several forms: changing a choice, narrowing the viable options, improving the quality of an execution, or setting a baseline against which the market can later be read. The common factor is not a headline ROI figure. It is a credible connection between the question asked, the evidence generated and the action taken.

Research earns influence before the readout, when it changes the question being asked.

Start with the decision, not the deliverable

The strongest practical change is to design each commission around a live decision rather than a method or output. “We need customer interviews” and “we need a segmentation” describe activities. They do not establish what the organisation will do differently once the work is complete.

At the outset, researchers should be able to state five things plainly:

• the decision owner and the point at which a choice will be made;

• the options that remain genuinely open;

• the customer, market or commercial uncertainty preventing progress;

• the evidence already available; and

• the observable consequence of choosing one route over another.

This does not make every brief rigid. Exploratory research is valuable precisely because it can reveal that the original question was too narrow. But it gives the work an intended landing place. It also stops teams from confusing speed with usefulness: a fast answer to a question that has already been settled elsewhere is not strategically influential.

For insight leaders, this means spending more time at the framing stage. The question is not simply whether stakeholders have supplied a clear brief. It is whether the business still has a choice to make. If the preferred direction is already fixed, research may be better used to refine implementation, identify adoption barriers or establish measures for the market response than to simulate a decision that has effectively passed.

Build a decision trace, not an ROI theatre

A lightweight decision trace can provide more value than an elaborate attribution model that nobody maintains. It should be recorded at three points.

First, document the starting position: the assumptions, options and known evidence before the study begins. Second, capture the decision: what was agreed, which findings were used and what was ruled out or amended. Third, revisit the decision later to see whether the expected customer or market signal emerged.

The final stage will not always produce a clean answer. A revised message may go into a campaign alongside a changed offer and a different media plan. A new service design may be launched while staffing, pricing and product availability are also shifting. That is normal. The aim is not to create false certainty, but to preserve the logic of the decision and improve it next time.

Over time, these traces can reveal where insight makes its most distinctive contribution. A team may find it is especially effective in identifying unmet needs early, challenging internally popular propositions, clarifying buying-group dynamics or translating behavioural data into a usable customer story. That is far more useful than a generic claim that “research works”.

Make accumulated evidence easier to use

The new study also shows broad use of AI for tasks such as analysing open-ended responses and synthesising past findings, while most practitioners still want human review. That pattern matters less as a technology story than as an operational one.

Insight teams already hold evidence that could improve a current decision, but it is often fragmented across decks, trackers, agency reports and the memories of people who have moved roles. A searchable repository can help, provided it retains the context that makes a finding trustworthy: who was studied, when, by which method, for what decision and with what limitations.

The commercial benefit is not merely faster retrieval. It is better research design. Teams can distinguish between a question the organisation has genuinely not answered and one where existing evidence is being ignored because it is difficult to find.

The profession’s challenge is therefore not to prove that every study caused a number on a spreadsheet. It is to show, with enough discipline to be credible, how customer and market evidence improved a decision while there was still time for it to matter.