Is your agency a provider, or is it a partner?
Editor’s note: Jodie Shaw is the head of global marketing at Kadence International, where she leads global brand, content and demand-generation initiatives for the company’s market research and insights practice. With more than two decades of experience in B2B marketing, she has built a reputation for turning complex services and emerging categories into compelling market propositions. Find Shaw on LinkedIn.
By the time most market research is commissioned, a decision has already been made. The target audience is defined. The category is accepted as attractive and the direction of travel has been aligned across the business. The brief formalizes that alignment, specifying who to speak to, what to test and how success will be measured before any external input is introduced. At that point, market research is not shaping the decision; it is just confirming it.
Take a standard consumer packaged goods launch into the “better-for-you” snacking space. Growth signals are visible, margins look acceptable and competitors are already there. The internal story locks quickly, and health-conscious Millennials become the target. The product is framed around convenience and premium cues. A quantitative concept test is commissioned to confirm appeal, purchase intent and price tolerance before launch.
It looks complete because nothing inside it is allowed to move. Everyone in brand or product roles knows that most new product launches fail. Estimates vary, but multiple industry analyses put the failure rate at 70-90% in the first year. Not because brands cannot execute, but mostly because the initial assumptions do not hold under real purchase conditions.
Despite the value of research, what happens time and again is the agency steps into a closed system. It can measure what it is given, it can refine within the lines, but it cannot ask whether the lines were drawn in the right place. That boundary is not subtle, and it defines the role before the work begins.
The moment the work tries to change direction
Something concerning shows up the moment the data refuses to behave. On the surface, the concept performs. The language fits the category, and respondents understand what is being offered.
When the product is placed in a real purchase context, the response changes. Interest softens under price. More importantly, the strongest signals come from lighter users of the category, while the brand's core buyers, who it expects to convert, show less urgency. The product might fit the category, but it does not displace what those buyers already purchase. Research shows it is acceptable but not needed.
This gap between acceptance and necessity is where most commercial failures sit. Research from NielsenIQ consistently shows that products can achieve strong top-box scores while failing to generate repeat purchase. The distinction is simple but often ignored. Being liked is not the same as being chosen, and categories are built on habit, not curiosity. If a product does not disrupt an existing routine or earn a place within it, it does not scale.
More often than not, the market research agency presents this clearly. The recommendation is not to abandon the concept, but to reconsider its structure. Narrow the audience to those who show real movement under constraint. In other words, re-examine the role the product is expected to play in the category before committing to scale.
But the implication is more uncomfortable than that. The opportunity, as currently defined, may not exist at scale. What looks like a mass-market proposition begins to fragment, revealing a narrower, less profitable segment that behaves differently than expected. This is the moment when research stops being supportive and becomes diagnostic. It identifies not just how to win, but whether there is anything meaningful to win at all.
This is the point where the decision should change. It would require reopening the brief, redefining the audience, adjusting the method and extending the timeline to accommodate further testing. It would also require unwinding internal alignment that has already been built around the original direction. The work would move from validating a decision to questioning it.
In practice, this is the last point where research can still change the outcome. Once a concept moves into validation, the range of acceptable answers narrows quickly. Methods are selected to confirm but not challenge. Audiences are defined to fit the proposition but not test its limits. By the time fieldwork begins, the decision is already directionally set. What follows is not exploration; it is calibration.
According to Bain & Company, fewer than one in 10 companies systematically revisit core assumptions once a strategic direction has been agreed internally. Alignment hardens into inertia. The cost of being wrong is deferred, while the cost of reopening the question is immediate. At that point, research is no longer being used to reduce risk; it is being used to manage internal confidence.
When the agency continues anyway
Often, the decision is taken to proceed within the existing plan. The same audience is carried forward. The same structure is maintained. Softer signals are acknowledged, then reframed as execution issues. Messaging will fix it. Positioning will close the gap. The market just needs time.
The agency continues the work on that basis. At this point, the distinction between provider and partner is no longer conceptual. The agency has identified a break in the opportunity's logic and recommended a change in direction. The decision has been made to proceed without making that change. The work that follows operates within that constraint.
This is not a failure of communication or analysis. The signal was clear, and the recommendation was made. The outcome is defined by the decision not to act.
Ignoring what the data was saying
Once the frame holds, the range of possible outcomes narrows. The research continues to produce valid insights, but those insights are bounded by what the work is allowed to question. The audience remains fixed, the product's role remains unchanged and the assumptions that shaped the brief carry through into execution.
The product launches with sufficient distribution to generate trial. Early movement aligns with expectations, supported by availability and initial marketing investment. The category is familiar, the proposition is understood and there is enough curiosity to drive the first purchase.
What does not follow is repeat behavior at the level required to sustain momentum. The same condition that appeared in the research holds in the market. The product is tried but not adopted. It does not establish a place in routine purchasing, and without that, velocity declines.
This is not a surprise outcome. It is the market expressing the same constraint that was already visible when the concept was tested under real conditions. The decision to hold the audience and the structure fixed set that path early. The data did not fail; it was ignored where it mattered.
This pattern is well documented. The Ehrenberg-Bass Institute has long shown that growth comes from penetration, not persuasion. Light buyers matter more than loyalists in driving scale. Yet many briefs remain anchored to heavy users and to the assumption of brand switching. When market research surfaces a different reality, it often conflicts with how the business believes growth should work. The data challenges not just the product, but the model behind it. And ignoring that signal is not a tactical decision, it becomes a strategic one.
What the agency sees but cannot change
From the market research agency side, this pattern is familiar. The signal rarely arrives as a single decisive metric. It shows up in shifts under constraint, in who responds with urgency and who does not, and in the gap between what people say and what they do when trade-offs are introduced.
These signals are not hidden. They are included in outputs, discussed in debriefs and documented in the work. The issue is not whether they are seen. It is whether they are allowed to change direction. When they are not, they are redirected into messaging and execution, rather than used to challenge the underlying decision.
The agency can point to the constraint, but it cannot remove it. This is where the definition of partnership quietly breaks. A partner is expected to influence direction, not just execution. But influence requires permission. Without it, the agency’s role becomes performative. Sure, it delivers insight, signals risk and documents alternatives, but the outcome remains unchanged. Over time, these conditions affect both sides. The client expects validation, and the agency optimizes within constraints. The language of partnership remains, but the behavior aligns with procurement.
When the decision has already been made
There is a tendency to assume the agency should push harder. But in practice, the constraint is not merely a matter of willingness. Pushing beyond the brief requires the client to revisit decisions that are already tied to budgets, timelines and stakeholder alignment.
It introduces delay and uncertainty at a point where most organizations are set up to move forward, not backward.
There is also a commercial boundary. The agency is contracted against a defined scope. Expanding beyond it risks delivery, complicates execution and strains the relationship. The work exists inside that reality as much as inside the data.
So many times, the agency adjusts. It continues to surface what it sees within the structure provided, knowing that the decision has already been taken. The role shifts from challenging the direction to supporting it.
Where responsibility actually sits
Responsibility does not sit in the method or the analysis. The research did what it was supposed to do. The limitation lies in the decisions that define what the research is allowed to question in the first place.
The client sets the brief. It defines the audience, the scope and the boundaries of the work. It determines whether those elements can be revisited or treated as fixed. And the agency operates within that structure.
This dynamic shapes how research functions in practice. It becomes a tool to reduce uncertainty within a chosen direction, rather than a mechanism to test whether that direction should hold. If the brief is closed, research reduces uncertainty within a chosen path. It does not test whether the path itself should hold.
If you want a partner, the brief has to change
If a brand wants a partner, the work structure has to change. Research cannot sit at the end of the process and still influence the outcome. It has to enter at the point where the audience, the proposition and the success criteria are still in play. That means testing demand before positioning, behavior before messaging and substitution before stated intent. Without that shift, the work will continue to produce insight, but not impact.
That requires a different starting point. The work begins before those elements are fixed, and it is allowed to reshape them. The questions are not limited to validation; they also address whether the opportunity holds up under real conditions.
This changes what the work produces. The initial opportunity may become smaller, more specific and more conditional. Some directions may not even proceed at all. The process becomes less linear and less predictable. It also avoids committing to decisions that the data shows will not hold under real conditions.
If the brief does not allow for that, the outcome is already defined. The agency will deliver strong work within the structure provided. The decision will proceed as planned. And the distinction between provider and partner will remain a matter of language, not behavior.
The distinction matters more now than it did a decade ago. Markets are less forgiving, and the cost of getting it wrong is higher. Distribution no longer guarantees success, and an early trial is easier to generate than sustained demand. The margin for error has narrowed.
Market research cannot sit at the end of the process and still claim to reduce risk. If it is not allowed to change direction, it is not de-risking the decision – it is just documenting it.
That is the line most researchers recognize, even if it is rarely said out loud.