A few days ago I got an email from Netflix telling me that “recently you may have had trouble instantly watching TV episodes or movies due to technical issues.”As a matter of fact, I did not experience any “technical difficulties.” I simply started and stopped a number of stream-able movies. So I skipped Netflix’s offer of credit to my account but started thinking about the future of business in the era of Big Data.
Twenty years ago, Mike Hammer published his landmark HBR essay titled “Reengineering Work: Don’t Automate, Obliterate.” It launched the business process reengineering movement that helped business executives think holistically about their companies and obliterate islands of data held captive by business functions and departments. With the help of IT vendors, reengineering also resulted in the creation of new islands of data, this time held captive by a specific process, served by a specific software application.
Cloud Computing (see here and here) promises to obliterate the barriers set up by applications and facilitate the sharing and mashup of data sets across different business activities. The set of technologies represented by the emerging concept of “Big Data” (see here and here) promises to facilitate the analysis of these data sets, so businesses can re-structure to become Web-driven businesses.
Hammer warned against a narrow definition of “automation” which resulted in simply adding information technology to existing processes. Similarly, a narrow definition of Cloud Computing will result in simply migrating existing applications to the cloud. And a narrow definition of Big Data will result in simply analyzing data that lives on an island without understanding its relations to other elements of the business.
Web-driven businesses will have a comprehensive view of Big Data in the Cloud and will constantly monitor, measure, and model the web of relationships between all elements of the business. But they will not just collect data. With an “analysis engine” fueled by the right models (algorithms), they will be able to understand the constant changes in their external and internal environments and adjust the business on the fly. Eventually, their knowledge of their business could reach a level that will allow their analysis engine to proactively predict future developments and simulate the possible outcomes of different responses.
Smart businesses have done that before, some with only small IT and thin analysis. A great example, thoroughly documented by Andrew McAfee is the clothing retailer Zara.
McAfee: “Zara is obsessed with making good decisions about what clothes to stock, but has configured itself so that people making these decisions operate in what looks like a ‘data vacuum’ – a lack of aggregated, filtered, and massaged information from throughout the corporation. This is because the good information that Zara’s business model requires is not the kind that’s easy to digitally encode, transmit, aggregate, and analyze. Instead, it’s information that comes from watching, talking, and listening, then using the computer between our ears to pattern match, draw conclusions, and peer just a little bit into the future.”
I will be the first one to argue that the “computer between our ears” will continue to be the most important business tool for the foreseeable future. But the watching and listening (and at one point, talking) part and assisting with the drawing conclusions part – all based on programs and models (algorithms) created by humans – will all be Web-enabled. And at a really Big Data level: Just imagine all the trillions of items P&G sells to 4.2 billion consumers in 180 countries, all connected to the Web, feeding data to local and global P&G analysis engines.
Which brings me back to Netflix. Shifting rapidly from a business based on atoms to a business driven by bits, they are already monitoring and measuring, proactively responding to potential customer dissatisfaction. But I’m sure they are also modeling, looking at customer behavior in the aggregate, trying to predict where their business will be in the future. And given that Netflix Instant accounts for 20% of peak U.S. bandwidth use, can you imagine what kind of analysis they do of Internet usage patterns, capacity constraints, and performance, everywhere in the U.S., and what kind of competitive and market differentiation this knowledge represents?