This article is part of an Essential Guide, our editor-selected collection of our best articles, videos and other content on this topic. Explore more in this guide:
2. - Big data and analytics: Let's go to the videotape: Read more in this section
- Daimler Trucks takes sensor data on the road
- AIIM: How to get big data off the hype curve
- HoneyBaked execs find the secret sauce in big data
- A data analytics strategy to improve business outcomes
Explore other sections in this guide:
- 1. - The premise and promise of big data analytics
- 3. - Getting down to business: Analytics in action
- 4. - All the analytics terms fit to print
A data analytics strategy for boosting business efficiency and revenueDate: Nov 04, 2013
Michael Koukounas, senior vice president of global scoring and analytics leader at Equifax Inc., offered valuable advice on how to build a strong data analytics strategy at the recent Kodak Alaris Global Directions 2013 conference.
In this Q&A with SearchCIO, he expands on how the consumer credit reporting agency is leveraging real-time data to improve its own business operations -- and to find new business. He also speaks to the importance of identifying the right data to increase operational efficiency and to the ongoing challenge of providing access to that data.
In your presentation, you gave a real-world view of data analytics in action at Equifax, including the six steps to building a strong data analytics strategy. Could you give a few highlights on what you shared with the audience and why data analytics is so critical to business success?
Michael Koukounas: I think a lot of people -- especially the data analysts -- get excited about data. They look at the data, and say, 'Wow, there has to be some answer there in that data,' and they go mining for that data.
It really is important to start with 'What is the problem? Where are the touch points that I can have an impact on that problem,' and understand the constraints of those touch points, whether it's systems constraints, whether they're regulatory constraints, or whether they're reputational constraints. There's going to be some constraints, and you have to work around those constraints.
How are you using consumer data as part of your data analytics strategy to solve problems for your customers?
Koukounas: Well, Equifax obviously has used consumer data to help banks decide who to lend to, who not to lend to, or how much to lend. But what is happening now with data is it's being used in a variety of different areas, for example, in the government.
We talked in the session about how there's fraud issues with the IRS -- people putting in fraudulent returns to get back returns, and billions of dollars are being lost in this. We're helping the IRS identify who those frauds might be, and stop those checks from going out.
We're also working with the healthcare services, the Centers for Medicare and Medicaid Services. With Medicaid [and] Medicare, and helping them identify, for example, 'Is Mike Koukounas a real person, and is he eligible for Medicaid services?' That information that Equifax has allows us to be able to help a variety of problems, way beyond the traditional credit decision.
With the day-to-day increase in cloud computing comes increased threats to security. What's your advice on how to best work with industry partners and business associates to ensure data security?
Figure out what you're trying to solve for. Figure out where you're trying to impact that, the constraints, and then choose the data you need.
Koukounas: I'm not a technologist. I'm not going to get into what the right infrastructure is, but it's also having the right procedures in place.
For example, with my analytical teams, I prefer they don't have access to personal and identifiable information unless they really need it for their analytics. There are ways around that, and it's not that I don't trust those folks, but why bother? Why have that information out there if you don't need that information?
So, putting the right procedures in place, even for your analytical teams, around what data is available and how to use that data is very, very important, and it allows for quicker, easier decision making.
We've heard a lot of talk at this conference about how big data is going to revolutionize the way organizations do business across industries. In your presentation, you specifically mentioned the importance of unstructured data, as well as structured data, for an effective data analytics strategy. Can you comment on this a bit more?
Koukounas: Well, it really is about making sure you pick the right data for the business decision you have. If unstructured data is going to be important to you, it's great to go out and get that unstructured data, but if it's not going to add a lot of lift and it's going to be onerous to get, you should consider not adding it.
It really depends on what you're trying to solve for, and that's really what I talk about in those six steps: Figure out what you're trying to solve for. Figure out where you're trying to impact that, the constraints, and then choose the data you need.
If it's more structure-oriented data, that's fine too, but again, you have to decide how you want to use that data and how important it is to making that decision in driving those [profit and loss]PNL changes.
Another big topic on the IT leader's plate is the unprecedented rate of technology change in the past five to 10 years. Looking at cloud, specifically, social computing, mobile -- how does your organization plan to keep pace?
Koukounas: Clearly, we're a company that has a lot of people coming to us and asking us, 'We have huge, vast amounts of data. Tell us how to monetize this data -- not only how to help make good decisions using this data for our company, but is this a revenue stream for me? Can I use this data to, you know, sell this data to other people, is it a revenue stream?'
Now, it's a balancing act as always, because that information is about your customers -- it's about you and how much of that information you want out there. It can be a competitive disadvantage, and so, we help companies understand in this new area where lots of information is made available, what can you do, what's out there, what are the tradeoffs, and how much money can you make using that data both internally or externally as a revenue stream.
Are there areas of enterprise technology that you think could work a lot better than they do today?
Koukounas: Getting my team access to data has never been easy, and you would think that in a data company it would be easy. Getting them that access -- making it easier for them to get access to data, but doing it in a way that's secure and doesn't create downstream issues for the company, or create reputational issues for the company -- has been a challenge. We'll continue to refine that over time.