This a reprint of a post from the MassTLC Unconference session "Generative Research: Small Data and Human Scale Insights" moderated by Jason Roberston, Senior Strategist, Continuum. It was one of the more informative sessions I attended, so I though I'd give it another airing here because I find the informational universe inundated with "big data" talk that seems overwhelming and not especially relevant to the early stage companies I work with most often. Rich insight can be gained at the "human scale" long before massive data sets come into play in a meaningful way for most emerging growth companies.

I just spent a two days at a conference in which the concept of Big Data and how to get the most out of it was discussed rather thoroughly. So, it was interesting to hear a pitch for a MassTLC Innovation unConference session on “small data” or human-scale research.

Generative research implies using research to generate ideas. The small data, or human scale, terminology indicates a method that requires close observation and deep analysis of a limited number of human subjects. Where big data is great for seeing trends and both historical and predicative analysis, small data focuses at getting at the “why” behind individual choices and understanding the human context and emotional backdrop that informs those choices. Generative research delivers:

Nuance

Why (not just what)
Contradictions and unspoken needs

C.D.E

Context (human level context)
Depth of understanding
Emotion

Flexibility

Every problem is different
You can change your method on the fly based on any number of inputs

Through close observation of the daily reality of a target audience, companies and organization can develop use cases that are reflective of an actual reality, not a marketing department’s educated guesses. Several examples were cited of inaccurate assumptions made by an organization about its target audience that were only uncovered through human-scale research. Among these was the example of a juice manufacturer that was targeting Mexican families. The product (a juice powder that needs to be mixed) was designed to appeal to children under the guidance of a parent. Ideally the juice would be mixed in the kitchen of the family home. However, close observation revealed that typical Mexican kitchens are the exclusive domain of the mother – children are rarely allowed in – which meant a fundamental disconnect in the positioning of the product that would have gone unseen without human-scale research.

One of the more interesting elements of the session was a group conversation around how to eliminate personal bias when questioning or observing a research subject. Some hints included:

Write all your opinions and bias out on paper or white board in advance. This often has the effect of defusing a simmering bias
Work with a team on research methodology extensively first, before going into the field. Acknowledge the bias and solve for it.
Let the subject lead the research – open ended questions are perfect here. Follow the subjects lead through her answers.
Be mindful of your personal presentation: If you’re interviewing people in lower socio-economic strata wearing a fancy watch and reading your questions off of the latest iPad may be a bad idea. Similarly, if your interview subject is a Wall Street titan, dress and act in a way that meets your subject on her or his own terms.

Human scale research enables a new perspective on data that:

Makes strategy better
Makes companies “smarter”
Informs better, more focused human-oriented product design
Allows companies to care about their audiences as people, not data points
Helps identify real world problems

Ultimately, small data should be used in conjunction with big data methods. It is not a substitute or in opposition to any of the large-scale research and analysis tools. These methods support each other and, taken together, give any organization the best chance a truly understanding its market and audience.

Small Data and Human-Scale Research