My colleagues and I were buzzing yesterday about a recent Fast Company article describing how Twitter can predict when people will get sick. (Apparently a new algorithm designed by a team at the University of Rochester can effectively monitor the spread of illness via people’s tweets and predict illness up to 8 days in advance!) We agreed that this is pretty nifty, but our feeling of wonder was tinged with a hint of exasperation at how far data scientists are pushing “Big Data” to learn about people. What’s next?
We’ve been talking about that same subject – how can we use new, unstructured data sources to learn about customers/consumers – with Researchers from across the membership for the past 6 months. We’ve found a lot of interest in the opportunities presented by new data sources, but a lot of skepticism about their value to Research, given their limited scope (many are purely behavioral and don’t address “why”) as well as their failure to meet the reliability and validity criteria of traditional Research tools.
We’ve also talked to some forward-looking functions that have recognized the negative impact that the new multi-source information environment has on decision-making. These Researchers are using new sources not to deepen their insights, but to provide the business with a coherent customer understanding that makes sense of different data from a variety of sources and eases decision-making. MREB members, learn more about this research.
While we’re on the topic, we have recently launched a new diagnostic to help Researchers determine the most valuable sources and methods to use in the Research toolkit. This 10-minute survey designed to be taken by individual researchers will provide respondents with a custom report identifying their biggest opportunity for sources and methods – those that are most valuable and underused. MREB members, learn more and take the survey here.