In the many training courses I run about extracting and narrating the story hidden in data, insights emerge as a crucial focal point. One striking observation I’ve made is that there’s no universal definition of ‘insight’.
When I ask attendees, many of whom have ‘insight’ in their job titles, about their organizational definition of the term, the responses vary widely. This diversity is a reminder to be wary of anyone pitching a singular ‘correct definition’ for an insight.
Common Descriptions of Insight
- A new perspective.
- An unrecognised truth.
- A change in viewing things.
- An understanding of underlying motivations or behaviours.
- The uncovering of the ‘why’.
- A gateway to new outcomes.
- Actionable, relevant, and often obvious in hindsight.
- Starbucks: Their success isn’t just about coffee; it’s about the experience. Howard Schultz envisioned this after experiencing coffee bars in Milan.
- Dove: Recognizing that only 2% of women view themselves as beautiful, they spearheaded the impactful ‘Real Beauty’ campaign.
- Apple: They discerned a significant consumer group that prioritized style and simplicity over multi-functionality.
- Betty Crocker: In the 1950s, an initial cake mix requiring only water was a flop. Research showed housewives felt they weren’t genuinely baking, leading to the reintroduction of eggs in the mix, making it a hit.
Research Findings vs. Insights:
All research should provide findings, but not all research yields insights.
Consider an example where we test ten TV ad concepts. If they all fare well, and our findings align with expectations, then the research generates valuable findings, but not necessarily insights.
However, if all the concepts perform poorly, we might find that deeper analysis reveals that the ads resonate only with heavy product users due to the exclusive language. Realizing that language is the barrier to wider success would constitute an insight. Our picture of the world has changed, and we can generate new outcomes.
When Are Insights Most Likely?:
Based on my 45 years in market research, I have discovered that insights predominantly emerge from projects linked to problems or failures. When research validates our expectations, it often yields valuable advice based on findings. But when faced with unexpected problems, we delve deeper, making the discovery of insights more probable.
When we are conducting research, we should not get too hung up on ‘insights’. We should always be able to provide useful, evidence-based advice. But we won’t always (or even usually) be generating ‘insights’. Or perhaps, we should re-cast the word insight to use it in the way data science uses, broadly referring to any useful finding from the analysis.
Want to learn more about storyfinding and storytelling? Check out the courses I am offering by clicking here.