Michael Li is founder and CEO of The Data Incubator, an eight-week educational fellowship founded in 2012 that prepares students with Master’s degrees and PhDs for careers in big data and data science. The program also offers corporate training to Fortune 500 clients.
Michael previously headed monetization data science at Foursquare and has worked at Google, Andreessen Horowitz, J.P. Morgan, and D.E. Shaw. He is a regular contributor to Harvard Business Review, O’Reilly, Wall Street Journal, Entrepreneur, Venture Beat, Tech Crunch, and Fast Company. Michael was a postdoc at Cornell Tech, earned his PhD at Princeton, and was a Marshall Scholar in Cambridge.
We recently asked Michael about data-driven marketing and the value of advanced data science training within organizations. Here’s what he shared:
Why is it so critical that organizations have trained data scientists today? What role does data play in the ability of an organization to be successful?
Harnessing the full potential of data requires developing an organization-wide data science and big data strategy. Strategies like this are common in marketing and eCommerce where there is a strong need to segment customers, combat fraud, identify upsell opportunities and analyze transactional data. For example, retailers like Amazon analyze the shopping habits and preferences of their customers with powerful predictive algorithms to tailor marketing strategies. But the possibilities go well beyond just retail into almost every industry.
What skills and tools should data scientists focus on honing today, specifically as it relates to marketing?
Our data science fellows that we place at advertising and marketing technology companies like Adtheorent or AppNexus are trained heavily in developing predictive analytics models in R and Python. These predictive models help these companies identify which customer segments are most likely to respond to particular campaigns, which are at risk of attrition and can be re-engaged with additional marketing or discount opportunities. They also help identify the best products to market to fine-grained customer micro-segments.
Our data science fellows are also trained in techniques like clustering, which leverage powerful machine learning and AI tools to identify novel segments of customers that go beyond the traditional demographic breakdowns. For example, instead of just targeting a segment for “mothers of young children,” you can use AI to find “childless customers who buy kids gifts for friends and family,” opening up previously untapped customer segments. Even basic exploratory data analysis and visualization tools help to uncover relationships in data and direct further analysis that are not intuitively obvious. A good data scientist works closely with marketing to fully understand the key problem the marketer is trying to solve so that together we can pinpoint which aspect of customer behavior needs to be better understood.
How can data-driven marketing drive the success of a Marketing Technologist?
Data-driven marketing is the future of how products will be advertised and sold. For example, a marketing technologist can ask a data scientist on their team what search copy works best for basketball fans during the weekend on a post-game day. By quickly digging into available data, a data scientist can understand what time of day is best for selling any number of products to segmented audiences.
What predictions do you have about the role of Marketing Technologists within organizations moving into the future? What impact will they make on the success of organizations?
The reason data scientists can understand these valuable tidbits is because they can access every ad the company has ever run and every interaction for every client. They can then help a marketing technologist connect that performance data to publicly available data, from Google Trends to ComScore to Rentrak. From there, they can create models to analyze and predict consumer behavior — and drive better creative and marketing decisions.
Marketing technologists can essentially eliminate a lot of time-consuming testing and debate. They can also empower marketing decisions by providing clarity and confidence through modeling. A team that can rely effectively on marketing technologists to help steer product development and advertising will be able to create campaigns that instantly resonate with their target markets.
What should organizations be doing today to prepare for how they’ll use data in the future? How can they ensure their employees are more fluent in data in order to be ready for their roles down the road?
Great data science is built on great engineering. Companies need to empower marketing teams with engineering support to enable marketing technologists and data scientists to do the best job possible. If data scientists are mostly doing very high-level aggregate analyses, they shouldn’t have to be writing low-level MapReduce each time. Slow data clusters make it hard for data scientists to iterate quickly and saps both their motivation and creativity.
Data science is a quickly evolving field so a company that invests in the education of its marketing technologists, engineers and data scientists will see returns from that investment for years to come. By paying for corporate training classes, managers can encourage team members to learn new tricks of the trade. Providing space to host meet-ups with academics and data scientists from other companies is also a great way to encourage continuous learning.
What trends or innovations in data science are you following today? Why do they interest you?
Two of the hottest trends in big data and data science are Spark (and related distributed computing technologies), which allow us to quickly process terabytes of data on cheap commodity hardware, and Neural Networks, which allow computers to “understand” unstructured data like images, video, text and voice data. Demand for our corporate training courses on these topics has more than doubled in the last 12 months and we expect to see it continue to increase.
Download the Ultimate Guide to Driving Revenue with a B2B Customer Data Platform to learn more about how to leverage your data effectively today!