Many aspects of our lives – such as the TV shows we watch on Hulu, search terms we enter on Amazon, and places we go with our smartphones – are being recorded, tracked and analyzed. The data sets that these and many other activities create are larger and more complex than anything we’ve dealt with before. Consequently, traditional data management and analytical tools are not up to the task or making the best use of this new tidal wave of information. The term “big data” has come to represent the technologies, software tools, methodologies, and even the organization structures, that are designed to manage and analyze these new massive and complex data sets. Proponents argue that companies that master the challenges of big data will crush their competitors that don’t.
Russell Walker’s From Big Data to Big Profits
Russell Walker is an Associate Professor at Northwestern’s Kellogg School of Management. His job at Kellogg is to develop and teach courses that give MBAs a deep appreciation for big data and what it can do for an organization. I recently read Walker’s book, From Big Data to Big Profits: Success with Data and Analytics.
If you want great examples of how companies like Netflix, LinkedIn and Amazon have been extremely successful at exploiting big data, then this book is for you. Although it occasionally mentions a few big data software tools, this book isn’t going to teach you how to get your hands dirty with Hadoop or MongoDB.
From Big Profits to Big Data is well-organized and gives detailed explanations of several big data concepts. The real world examples consistently look at the critical dimensions of big data: velocity, precision, variety, and scale.
Walker advocates using big data to identify and exploit multi-sided business models. That means using big data to create new offerings, and when those offerings create more big data, you can use that data to create even more offerings. Throughout the book, Walker discusses inverting data, which means taking data originally collected for one purpose and using it exploit a new opportunity. For example, a cellular phone company could use the massive amounts of location-based user data that it collects to determine the best ways to invest in building out its mobile network. You could also re-purpose (i.e., invert) that same location-based data to determine, for example, the volume of people who pass a specific billboard over time. So by using the cellular phone data to observe the billboard, you could make a far better estimate of how much you should pay to advertise on that billboard. Moreover, mobile phone data contains a great deal of additional information about the people who pass the sign, or as the book would say, there is a high degree of “precision” in the data. Consequently, you can use that additional precision to improve your estimate of the billboard’s advertising value.
Walker argues convincingly that a centralized group should usually handle the data science function rather than locating teams within individual lines of business. A key advantage is that a centralized data science group is better able to combine data from multiple business units across the enterprise to exploit new opportunities, or put another way, invert the data to create more sides in a multi-sided business model. Walker also suggests that organizations should consider establishing leadership positions such as Chief Data Officer and Chief Data Scientist.
In the final chapter, the book describes a SIGMA (sources of data, innovations, growth mindset, market opportunities and analysis) framework for scoring an organization’s big data efforts. Providing a practical framework was a great way to end the book because it closes the loop on the concepts presented in the preceding chapters.
Big Data in Practice
From Big Data to Big Profits goes into detail about many companies that have proven their big data chops, like LinkedIn, Amazon, and Netflix. I was particularly intrigued with the Netflix example. It would be difficult not to talk about big data strategies while telling the story of how Netflix beat a household name like Blockbuster and then rose to become the 800-pound gorilla of paid on-line video streaming.
The name of the game for on-line subscription services like Netflix is to increase conversions and decrease cancellations. I’ll take a guess and say that the folks at Netflix first fell in love with analytics when they ran an A/B test demonstrating that they could double their subscriber conversion rate by putting “Cancel anytime. No questions. No hassles.” on their credit card entry page. But as the book describes, Netflix is also clearly harnessing the power of big data to increase retention (i.e., reduce subscription cancellations) by providing a better service that Netflix tailors to individual users. Netflix has invested heavily in creating a recommendation engine that does a great job of predicting what video a subscriber would like to watch. And they do that by exploiting big data extraordinarily well.
Based on my experience in the on-line media industry, I believe that Netflix is using big data for another purpose, too. Because big data can help it gain insights about their customers, how they interact with content, and how long those customers are likely to continue their subscriptions, it is very likely that Netflix can estimate how much a specific movie or TV show is worth to them in dollar terms. As one content rights holder told me: “Netflix gives me a list of the titles from my catalog that they want and lets me know how much they’re willing to pay. They don’t take everything. I’m not sure how they make their picks or set their prices.” Clearly, big data is playing a role informing Netflix what content it should acquire and how much it’s worth to them. In the situation where Netflix is negotiating to purchase content rights, big data gives them an asymmetric information advantage over the content rights holders. And I’m fairly confident that big data also plays a significant role when Netflix determines how much it should invest in original content.
As all the examples presented in the book show, big data can yield far more benefits than marginal tactical advantages to the companies that master it. Indeed, effective use of big data can also give organizations a massive strategic advantage that may enable them to demolish their competitors.
The Final Word on From Big Data to Big Profits
From Big Data to Big Profits is an excellent big data primer for management level folks. After reading the book, you’ll have a comprehensive understanding of how some of the best companies use big data today and a useful framework for evaluating your organization’s big data efforts. The author does an excellent job of explaining key concepts and tying everything together at the end. From Big Data to Big Profits won’t make you a data scientist, but you’ll at least be able to talk intelligently about big data and discuss the appropriate strategies for your organization.
See From Big Data to Big Profits: Success with Data and Analytics on Amazon. The book’s author also runs a two-day big data course at Northwestern University.