Most communicators develop research-backed stories backwards: they brainstorm questions, field a survey and hope the data reveals a story. Nathan Richter, senior partner at Wakefield Research, says that approach turns research into "a fishing expedition, leaving plenty of data but no clear takeaway or story."

After more than 20 years developing research-driven thought leadership, Richter argues the strongest studies for earned media begin with a hypothesis. Researchers should then shape the questionnaire around that hypothesis, rather than chasing a story after the research is complete.

What separates a topic from a testable hypothesis?

A hypothesis gives a broad subject clear direction. "Consumers are worried about the economy" is a topic, Richter writes. "Consumers are cutting back but continue spending on convenience" is a testable hypothesis, because it introduces a tension worth investigating.

The hypothesis should rest on evidence: existing research, reporting or observed changes in behavior. Richter offers a framework for structuring it:

• "We believe X is happening because Y"
• "And the consequence is Z"

Before writing the questionnaire, communicators should think through the headlines the research might support, why the story matters and which supporting findings would deepen the narrative. Who is most affected? What explains the behavior? What are the outcomes?

How do you translate a hypothesis into survey questions?

A question-first survey might ask whether consumers worry about prices or which purchases they have reduced. Richter says these are important questions, but they fail to show how behaviors relate, the emotion behind them or the urgency of the moment.

A hypothesis-driven questionnaire would examine whether the same people reporting cutbacks are maintaining spending on convenience, and why. It would measure financial concerns, changes in spending across categories and the reasons behind those choices.

The design should also allow comparisons among relevant groups: whether time-pressed parents make different trade-offs than retirees, or whether the pattern varies by income.

Each question needs a clear purpose: test the hypothesis, explore an explanation or establish useful context. Results that contradict the original hypothesis do not mean the research failed. Unexpected findings can reveal an emerging shift, challenge an assumption or point to a more compelling story. A hypothesis provides direction; the findings determine where the story goes.

How does hypothesis-driven research create more earned media?

If the goal is earned media, research should read like a story, not a report, according to Richter. Brands need a hook with urgency, a headline that spells out the main idea and supporting findings that add depth and context. The strongest PR-driven research gives journalists more than a single statistic to cite and a narrative to build on.

That narrative also multiplies opportunities for communicators. A strong headline grabs attention early, while supporting findings create additional angles, spark fresh questions and open follow-up opportunities. Because the story is built from the research itself, it stays grounded in data, not conjecture.

The difference between research that yields a few data points and research that yields a strong story starts long before the questionnaire goes into the field. Done correctly, Richter writes, the survey becomes a story — and that is how brands build real thought leadership.