On November 12th and 13th, 2014, Trexin joined more than 250 attendees at the Predictive Analytics Innovation Summit and the co-located Business Intelligence Innovation Summit at Chicago’s Swissôtel. Senior executives and thought leaders from companies like Walgreens, Netflix, bitly, Microsoft, USAA and Verizon presented their latest thinking on how to leverage predictive analytics and advanced business intelligence to deliver business value.
Among the many speakers, a few stood out as especially insightful:
- Dr. Bipin Chadha, Business/Enterprise Architect at USAA, gave a great talk about systems thinking as an approach to maximize the efficacy of analytics. He showed several insightful System Dynamics diagrams in the style of Peter Senge’s Fifth Discipline http://en.wikipedia.org/wiki/The_Fifth_Discipline
- Kaiser Fung, author of several well-known books on data analysis [Numbers Rule Your World; Numbersense], talked about accountability in Big Data. His central thesis, that there’s no evidence of increased accountability when using Big Data, resonates in an environment where Big Data is seen as a magic wand that will solve analytics problems that have proven difficult for decades.
- Kelly Uphoff, Director, Experimentation & Algorithms at Netflix, discussed quasi experimentation as a way around the common issue of not being able to A/B test certain hypotheses. The rigor and science she brought to the problem at hand was refreshing and showed how “perfect is the enemy of good” in this context.
- Colin White, Executive Director of Clinical Intelligence at Avera Health, spoke about data mining in healthcare. His comment that “doctors don’t want dashboards, they want prescriptions” can be seen as a warning against creating yet another dashboard that “looks like analytics” but in reality is essentially useless to its consumers.
- Finally, Saurabh Bhatnagar, Senior Data Scientist at Rent The Runway, gave a technical yet an entertaining talk on predictive modeling in retail – specifically, renting dresses online. His was the only presentation to include slides with both pictures of dresses and fairly complex mathematical equations. I also liked his comment to a questioner who was asking about a potential optimization of the algorithms used: “You’re thinking like an engineer, not a woman”. This reminder that all of the math and computer firepower of advanced analytics should relate back to human goals was a great way to close out the conference.
While Business Intelligence has been around a long time and is starting to mature, true Data Science and Predictive Analytics are just beginning to bear serious fruit, even in Fortune 500 companies. It’s well-known that there’s a tremendous amount of interest in using data to optimize business operations. While exciting in context of the “grand vision” for analytics, conferences like these – and the relatively small value achieved to date by many predictive analytics efforts – illustrate that practitioners and executives are beginning to understand the tremendous amount of work and skill needed to get maximum value from data. The future is bright for analytics, but there’s plenty of work to be done.