The Problem Isn't the Output

Most research that comes back wrong was delivered exactly as briefed.

The analyst read the ask, did the work, and sent it over. The problem started earlier, in the three-minute brief written between two other tasks.

"Can you pull some background on X." "I need a market overview for the IC." Instructions that feel specific enough in the moment but contain no information about what decision the research needs to support, what format is useful, what is already known, or what the deadline actually means in practice.

The analyst optimises for completeness. It's the only thing they can control when the brief is unclear. The output is thorough, well-formatted, and largely useless for the purpose.

What Happens Next Is the Real Problem

One cycle of rework costs a few hours. That's irritating but manageable.

The compounding effect is what matters.

When research consistently comes back needing adjustment, seniors lose confidence in the function. They stop sending the harder questions. They do a quick version themselves. Or they go into meetings underprepared because the brief-to-output cycle takes too long to be useful at pace.

The research function doesn't collapse. It quietly becomes something else: a filing service, a basic screen, an admin task. The capacity to generate genuine analytical input, the kind that actually shapes how a deal team thinks about a sector or a target, atrophies because no one is investing in it properly.

That shift rarely gets named. It just becomes normal. The team stops expecting much from research, and research delivers accordingly.

Where the Gap Actually Sits

The failure point is almost never the analyst's capability. It's the assumption that the brief and the actual need are the same thing.

They rarely are. The person commissioning the research has a mental model of what they want. That model is almost never fully articulated in what they write down. The analyst works from the written brief, not the mental model.

A good process creates a moment where those two things are aligned before the bulk of the work begins.

The Pre-List Check

When building a target or company list, rather than delivering the full output, the right move is to share a sample of 20 to 30 companies with the deal team first.

Almost every time, there is feedback. The geography was interpreted too broadly. A particular business model should be excluded. The revenue threshold was misread. There was something specific in the deal team's head that never made it into the written brief.

That feedback, gathered at sample stage, takes 20 minutes to incorporate. Gathered after 300 companies have been screened, it costs days and often triggers a rebuild from scratch.

The sample stage does something else too. It forces the deal team to articulate what they actually want, in response to something concrete. Abstract feedback on a brief is hard. Specific feedback on 30 real companies is easy. That conversation surfaces misalignment that would otherwise sit hidden until the full output lands.

The same logic applies beyond list-building. A one-page summary before the full sector note. A draft structure before the model. A quick verbal check before the written output.

The brief is not a document you write once. It is a hypothesis you validate before committing the work.

The Question Worth Asking

Think about the last piece of research your team commissioned.

Was there a checkpoint before the full output was built? And if there wasn't, how much of what came back actually changed how the team thought about the decision?

If the answer is "not much," the issue probably isn't the quality of your research function. It's what you're asking it to do, and how.