By Dave Hannon
I was having a chat with my three good friends Zooey, John, and Samuel L the other day and we all agreed the best thing about Apple's Siri is that it doesn't dispense advice on its own without you asking. Instead it provides you answers to specific questions. Like Zooey wanted to know where she could get soup delivered from--she's sooooo lazy--and Siri told her.
And it got me thinking about the value of big data and real-time analytics (that's when Zooey and John left the room--Samuel L will talk about anything though) and how when used correctly, a big data solution should answer questions. The conversation went something like this:
Dave: "I think some companies feel like you can just plug in SAP HANA or Hadoop and it will instantly start spitting out new business efficiencies for you. Like a magic eight ball."
Samuel L: "You know I was just thinking that the other day while making my gazpacho. Like the big data gods will strike down upon thee with great vengeance and furious anger those who attempt to poison and destroy.."
And that's when I left the room because 1. I hate it when Samuel L does that and 2. I had to go pick up Zooey's soup because she doesn't own a rain coat and the only place that delivers was flooded. Apparently Siri didn't know that.
As I was waiting in the deli for the soup, though, I read a column on the Wall Street Journal's CIO Journal page, where Bill Abott and Chris Curran of PwC phrased it this way: "The key to turning informational 'ore' into insightful gold is posing intelligent questions and formulating hypotheses, then digging for answers to prove or disprove theories ... We need to ask the right questions about the business before prematurely plunging into the data."
That's it. It's exactly what I was trying to articulate to Samuel L before he started on his rant.
But for me, having a real-world example always helps so I wanted to find out what kind of question a business might ask that data could solve. I was going to call John and ask him, but his questions in general are always very existential and, frankly, a little creepy sometimes. Instead, I poked around on the WSJ site a bit more and found a good example of a company that had a specific question to ask. Retail giant Sears needs to know what is the optimal price to charge for its products. All of them. And as Rachael King of WSJ's CIO Report describes, that basically amounts to a continuing game of trial-and-error. Set a price for a product, see how it sells, analyze that data, adjust if necessary, repeat.
"Because of the cost and the time involved in processing data on mainframes, it took Sears about eight weeks to run its price elasticity algorithms to find the right price for items. The retailer could only use about 10% of the data it collected because using more would have taken more time and cost more money."
With a specific question--how can we speed up the analysis of our pricing strategies--Sears successfully implemented a big data solution and started getting answers.
So they're out there! Companies asking questions that can be answered by big data. In fact, sitting in the deli thinking about it more, I remembered a recent insiderPROFILES story we did that provided another great example. Chipmaker NVIDIA had a specific question to answer as it transitioned 40-nanometer manufacturing technology for its products: How much of the new product would NVIDIA’s customers want, and how much of the existing sizes should the company continue to make? And how long would it take its customers to fully transition?
NVIDIA found the answer by building a customized interface on top of SAP APO for the new inventory forecasting solution, using SAP BusinessObjects Web Intelligence. You can read about it here.
Wow, I was really onto something here. In fact, I soon realized that my pals at the new site Data Informed had some similar examples of companies using big data to answer specific questions. In describing Hilti's presentation on SAP HANA at SAPPHIRE NOW, Michael Goldberg writes "By implementing HANA, Hilti is hoping to analyze some old questions and add layers of analytics to fashion entirely new insights. In one important case so far, it has succeeded in analyzing data in new ways: the company can detect historically, and simulate future actions, for changes in pricing and product mix on customer behavior and sales."
Boy that sounds similar to what Sears is doing. I was really nailing it now.
Unfortunately, I got so engrossed in my own thoughts, I had lost track of time. And by the time I delivered Zooey her soup, it was cold. And on a rainy day, Zooey can be a bit cranky. The conversation went something like this:
Zooey: "I've been dancing up a storm here and you bring me cold soup? Seriously? What am I supposed to do with this?"
Samuel L: "I'll have it. I love gazpacho!"
Dave: "Maybe you can ask Siri how to put a pot on the stove and pour soup into it?"
Zooey: "Put a what on a what?"
What I asked next isn't really important. But what IS important is the lesson you can learn from some of these examples: if you're like a lot of companies and thinking about what you need to do before implementing a new big data solution (and that includes SAP HANA) you need to know your implementation strategy and technology requirements and all that stuff. But you also need to have your list of "questions" ready. In other words, you should have some specific business challenges in mind when you implement to expedite the ROI and show business value.
I'm off to the doctor to get a check up on my orbital fracture. Zooey might not know how to heat up soup, but she does have an "answer" for smart aleck questions that explains why she keeps that sauce pan around.