Hadoop is no longer a science experiment — it’s transformational. That was the common thread among the keynote presentations at the Hadoop Summit held in San Jose, Calif. this week. Hortonworks President Herb Cunitz drove the point home over and over again. His customers provided the supporting arguments, discussing business wins that they had already experienced and charting the paths to what they seemed certain would be still bigger gains.
When management consultant Geoffrey Moore took the stage, he declared that Hadoop had crossed the chasm and grown beyond a disruptive technology to one that was primed for widespread adoption. If anyone is qualified to make a statement like this, it’s Moore. He wrote the book, "Crossing the Chasm," the bible that startups use to take their products mainstream.
Moore reflected back to 2012 when he had keynoted at the Hadoop Summit. The sentiment in the room at that time was “All we are saying is give Hadoop a chance,” he reflected.
The presenters he had heard at the conference this year, he said, were pointing to real world use cases. They came from enterprises like TrueCar, Symantec, Rogers, Schlumberger, Home Depot and Verizon to name a few.
#hadoopsummit Hortonworks does effective job of marshalling marquee custs for panel: Home Depot, Rogers, Schlumberger, Symantec, Verizon— Merv Adrian (@merv) June 11, 2015
Real World Use Cases
Russell Foltz-Smith, VP data platform at car buying site TrueCar said that his business bet big on Hadoop and that it won huge rewards as a result. You can’t be the brain of the automotive industry if you can’t process all the data, it seems. And there’s a lot of data.
Foltz-Smith said that TrueCar takes in 12,000 data feeds, 65 billion data points and manages about 700 million car images, each of them is vital. "If there is no vehicle image, the car doesn't exist (as far as the consumer is concerned),” he said. "And there is a ton of intelligence embedded in those images."
What’s impressive about TrueCar’s Hadoop success story is that it has 20 million buyer profiles and 600 terabytes of data in active use at any time. The prices and other information it provides customers are updated every 15 minutes. It’s a business differentiator, said Foltz-Smith.
David Lin, director of cloud engineering at Symantec, who sat on a customer panel at the conference, said that Hadoop has helped his company provide an even more intelligent service. When you put all of your data in one place, he explained, it comes together in unexpected ways. “Our analysts are seeing things they’ve never seen before,” he said. And discoveries using technologies like “Hadoop and friends” (the catch phrase used during the conference to describe the products in the greater ecosystem) are coming quickly — “in seconds,” said Lin.
“Time to protection” is a differentiator when your enemy is an attacker trying to do something “not good,” as Lin put it.
Rob Smith, executive director for IT at Verizon Wireless said that Hadoop plays an important role in helping his company map the customer journey. By collecting a wealth of data not only from internal sources but also streaming from outside the company, like from the web and social media, it is able to see the customer experience and bring it into calls when there are problems. The same goes for the proactive identification of what could become bad experiences.
What’s the point, aside from providing better customer experience? Preventing customer churn. It’s expensive to acquire new customers. Big data and Hadoop can make it cheaper to keep them.
Chris Dingle, senior director of audience solutions at Rogers Communications said that having customer information in a data lake helps his company get a 360 degree view of the customer. When data is stuck in silos, marketing might know something about you that customer service or accounting may not, so data assets are being underutilized and business advantages that could be realized are not.
When end users are given more information to make decisions, they can “show up like rock stars,” said Dingle.
Getting the right product, to the right customer, at the right time and price is the goal of almost any business that wants to close a deal with a potential customer. Having all the information needed to do so and being able to glean insights from it quickly is an entirely different story, yet it’s one Home Depot is eager to tell. Sam Gentsch, manager of IT at the company said that Home Depot can now keep 10 years of customer data to crunch, something that would have been difficult in the past because of expense.
Storing the data in commodity hardware saves money, but that’s not all. Hadoop gives Home Depot managers the power to dig into the data, to work faster and help the customer. “What more could you want?” asked Gentsch.
Not only that, but when you enable users who might not be best friends with statistics to access data, you sometimes see incredible results. “They look at data in such odd and creative ways,” said Gentsch.
It's Not About Tech, It's About Business
It’s worth asking why everyone isn’t doing Hadoop after hearing stories like this. While the panel stopped short of using Nike’s “Just do it,” catchphrase, that was the underlying message.
“There’s lots of uncertainty,” said Lin. “There are the 'it will never work' people, the insecure people. Kill the haters, and just dive in. Smart people, cool tech, they’ll figure it out.”
Other panelists suggested that you start with a use case. It could be that some of the companies who experimented with big data stopped in their tracks because they didn’t know what to do with it.
If the Hadoop Summit is any indicator, one thing that Hortonworks seems to have done especially well is to set its customers up to win by helping them define what a “win” might look like before they got started.
At the end of the day, it’s not about tech, but business.
“Science fair Hadoop is not what enterprises need,” we heard someone say.
Our takeaway from the summit: Quit selling Hadoop. Sell the business story and help your customers along the way.