TDWI | Training & Research | Business Intelligence, Analytics, Big Data, Data Warehousing

Welcome to Analytics 3.0. At TDWI's Las Vegas Conference, keynote speaker Tom Davenport argued that you don't have to be a Silicon Valley start-up to practice data-analytics innovation.

Data Analytics Innovation within Everyone's Reach

Welcome to Analytics 3.0. At TDWI's Las Vegas Conference, keynote speaker Tom Davenport argued that you don't have to be a Silicon Valley start-up to practice data-analytics innovation.

Welcome to Analytics 3.0. At TDWI's Las Vegas Conference, keynote speaker Tom Davenport argued that you don't have to be a Silicon Valley start-up to practice data-analytics innovation.

• By Stephen Swoyer
• February 23, 2015

At TDWI's World Conference in Las Vegas, keynote speaker Tom Davenport argued that you don't have to be a Silicon Valley start-up to practice data-analytics innovation.

Davenport is a rock-star in analytic circles: he's a senior advisor at Deloitte Analytics, a professor at Babson College, and a senior research fellow at MIT's Center for Digital Business. He's perhaps best known as the co-author of one of the most influential articles in recent analytic history, Data Scientist: the Sexiest Job of the 21^st Century, published by the Harvard Business Review in 2012. In his TDWI keynote, Davenport told attendees that sexy is what sexy does.

There's a tendency to think of big data analytics as the province of Silicon Valley start-ups and other scrappy players, but plenty of big, old, established companies -- from United Parcel Service (UPS) to consumer products powerhouse Proctor & Gamble (P&G) to agribusiness giant Monsanto -- are remaking themselves into data-analytics competitors.

As a case in point, Davenport cited UPS' efforts with “Project Orion,” a roughly $300 million effort he described as “the biggest analytics project on the face of the earth.” UPS didn't undertake Project Orion on a lark or as an experiment, he argued: it did so because it's been wrestling with small and big data issues for decades now.

Thanks to the availability of new technologies, concepts, and methods -- which, collectively, we call “big data” -- UPS and other organizations are finally able to do something about these issues. These companies -- and, yes, their scrappy Silicon Valley kith -- are exemplars of a shift to what Davenport called “Analytics 3.0.”

Don't think of Analytics 3.0 as a discrete new category, or as some kind of new paradigm that completely upends the status quo. Think of it, instead, as the next step in a kind of analytics maturity model.  In Analytics 3.0, companies are combining their traditional business intelligence (BI) and analytics practices (Analytics 1.0) with their investments in big data platforms and technologies (Analytics 2.0) “in a positive way that … takes advantage of the strengths of both.”

Established companies are already using analytics to accelerate certain aspects of decision-making, as well as to reduce the cost of decision-making and to improve the quality of decisions.

With Analytics 3.0, they're upping the ante, Davenport explained.

“The most sophisticated companies have been entering the [Analytics] 3.0 zone. [They're saying] we're using analytics, we still need to make our decisions better, but we're doing data products as well,” Davenport told attendees. The most intriguing new analytic use cases, he said, involve the creation of “data products” based on analytic insights.

“Mainstream organizations … over 100 years old … [are] saying we're not in the agricultural product business or the gas turbine business: we're in the data and analytics business. Yes, we do need to do decision support, we need to do that at a much greater scale than ever before, and we [also] need to do this data products stuff.”

This last is the specialty of the data scientist, and the data scientist isn't like the back-office analytics guy of yore. “A lot of these data scientists … are not people who want to be in the back office anymore; they want to be [involved in] making these important decisions or at least [on the bridge] right next to Captain Kirk making these important decisions in the front office.”

The data scientists driving the shift to Analytics 3.0 have very different expectations about the products they're creating and the speed at which they're creating them. “The expectations for speed of change … [are] very high in this world,” Davenport said. “[The data scientist says] 'I want to be creating product, I want to be creating features, demos at least.' Internal decision support is not as important to them. They have very high expectations and intolerance for bureaucracy.”

There's a popular conception that the data scientist is the “unicorn” of this era of IT. There might be something to this.  When he interviewed data scientists a few years ago, Davenport says, he found that a surprising number of them had backgrounds in experimental physics -- but what's most compelling about the shift to a data product economy is that everybody can play, he says.

“In the data economy, everybody can do this data products thing. … Every software company has the potential to say 'We can not only sell you software, we can tell you things that we've learned about our customers and their use of our software,” he said. “Imagine what SAP knows about how organizations operate themselves or what Salesforce knows about what companies sell.”

Quite aside from the data scientist, Analytics 3.0 requires new human talent in the form of big data champions who can articulate clear business objectives; data architects who can pull everything together; quants who can make sense of it all; storytelling “translators” who can make sense of the quants; product developers who are versed in building data products, not traditional products; and an entirely new class of user that Davenport dubs the “analytical amateur:” someone who can and will make effective use of analytic insights in doing her job.