Over 100 big data professionals gathered in Chicago in late June for Innovation Enterprise’s (IE) Big Data & Analytics Summit. While this was a relatively new event, IE was able to attract speakers and attendees from many of the largest retailers. It was a great change of pace from mega conferences with 10,000+ attendees like NRF’s Big Show or IRCE. We were able to have many detailed conversations about real problems in the big data trenches, primarily regarding marketing and customer analytics. Similar to what we learned at IRCE, everyone seems to be challenged with multi-channel attribution and engagement. A few notable highlights included:
Michaels: Christine Smiley, Director of Consumer Insights, touched on an often overlooked, but critically important subject in her presentation “Explaining Analytics Across your Organization”. She illustrated how describing outcomes with a “less is more” approach using images, analogies and less statistical results was a hard skill to master but necessary for articulating recommendations to CEOs and other C-level executives. Discovering statistical significance within a dataset is half the battle. Obtaining senior buy-in to implement a decision based on actionable data is the other half.
OfficeMax: SVP Rajeeve Kaul offered detailed insights about his experience and covered several model applications, methods and sample problems common for B2C retailers. According to Kaul, it is even more important for mature industries with thin margins to invest in big data tools and that even a small percentage change in revenue or cost savings can drive significant cash flow.
Sears Holdings: Peter Zhang, Director of Customer Analytics, discussed that while using data-driven commerce tools may allow you to predict components of customer lifetime value, the goal should be to predict how to acquire the most profitable customers. He advised attendees to start with the low hanging fruit to demonstrate the positive impact of analytics. We surprisingly learned that his group does not use Hadoop, unlike other groups at Sears that frequently tout the framework.
Whole Foods: Chris Taylor, Executive Coordinator of the Global Data Team shared an inspiring story about how big data, combined with a flat-company culture and an opportunistic real estate strategy allowed Whole Foods to make a bold investment to expand into Detroit. This decision took guts and ended a 15-year effort to develop this property as Detroit had lost 25% of its population in the last decade. She mentioned that Whole Foods has failed hundreds of times with incorrect decisions that were based on mining big data but the benefits of being right (e.g., Detroit) had materially outweighed the failures. However, having a culture that embraces risk-taking with calculated data-driven decisions best facilitates long-term retail innovation. She also made it clear that defining the decision that needs to be made is the most difficult part when extracting insights from big data in the grocery space. When determining causation vs. correlation, common sense to look past false positives is quite important.
Conclusion: Investing in big data platforms is in the early stages of adoption and large retailers with many channels and physical stores are challenged to harness their data to optimize marketing, pricing and their competitive position. According to a recent TechRepublic/ZDNet survey, 90% of companies have not invested in a big data platform with over 23% of respondents suffering from a lack of resources. Many at the conference were still using legacy SQL databases and analytics tools. Software that allows CMOs, CIOs and analysts to evaluate real-time data, provide easy-to-understand visualizations or had a suite of different predictive models were in high demand. All expressed an interest in testing machine learning, especially for applying adaptive intelligence for marketing and personalization.
We will definitely attend future IE events. We expect that IE will expand their retail-focused big data summits and add value by solving common retail problems such as on-site search tuning, price optimization and marketing attribution. These topics alone could justify a full day of speakers evangelizing how they would tackle each problem.
SwiftIQ provides technology to turn your company’s data into a platform to create intelligent, data-driven applications and predictive insights.