"Presenting Results of Your Analyses so That People Take Action”

Your team is exceptionally skilled at crafting incredibly clever and elegant analyses of complex problems using enormous quantities of data, but then comes the hard part - getting your customer (or your manager) to take appropriate action based on the results of your hard work. While sometimes this happens, often it does not, and this can be very frustrating. The good news is that with the right techniques, you can reduce the frequency with which this happens to you.

As people highly trained in technical disciplines, we tend to incorrectly assume that people make decisions based strictly on logic and reason. However, frequently they do not, and how information is presented has a significant but invisible impact on how people interpret that information and the decisions they make based on it.

For example, imagine that there are two groups of investment professionals. The first group is told that a sophisticated algorithm parsing petabytes of data has been used to estimate that there is approximately a 90% chance that Oracle stock will increase by at least 8% by the end of this year. The other group is told that the same algorithm has estimated a 10% chance it will not. Recognizing this is the exact same information expressed in two different ways, you might intuitively believe that the two groups of investment managers would make more or less the same choices about whether or not to buy Oracle stock. But in fact, they make very different choices.

In this discussion, we will explore several surprising ways in which decisions are influenced by which choices are presented to them and how those choices are described, and what the implications are for product teams and data scientists in designing your services and presenting the results of your work.