AutoData: A New Methodology to Answer Tough Questions

Digital disruption is impacting every part of the economy, including MIT CISR research! In this project, we are experimenting with a new methodology we call AutoData. In AutoData research, we embed scripts into a company’s systems—with their permission—to collect data automatically, and use clever analytics to answer key questions. We use this real-time data/analytics combination to answer previously hard-to-pin-down questions like “what does it take to sell effectively on a mobile device?” and “what does the customer say vs. what they do?” and “how do we predict when we could lose a good customer?” We will start with “what do customers say vs. what they do?”

AutoData is information we get directly from systems that we typically can’t get any other way. To answer the initial study question, we will look at data such as differences in NPS across segments, engagement measured by time spent in channels, value of the customer, customer interactions, customer loss, etc. The short-term goal is to answer the question of whether there are significant differences between what customers say and what they do. The long-term goal is to show how to answer critical questions for a company with automatic data collection and clever analytics in a short time—i.e., AutoData. We are actively looking for sponsors to volunteer to participate in this research; please contact Stephanie Woerner if you’re interested.

Research questions include:
(*) Is there a difference between what customers say (in feedback surveys like NPS) and what they do (in buying/using products or engaging with the company)?
(*) Can we demonstrate that AutoData research works to answer important questions?
(*) What types of management questions could be answered using AutoData?