Research

Podcast: Digital management solutions are basic competitive necessity

Digital universe. Exabytes. Data fluidity value. Master data management solutions. This is the language of the future of business. As the amount of data companies attain and store grows, so too must the ability to deal effectively with this digital avalanche. Michael Goul is a professor of information systems at the W. P. Carey School of Business. Here, he discusses how businesses will have to learn to manage unprecedented amounts of data as a means of gaining a competitive edge.

2008-05-21

Digital universe. Exabytes. Data fluidity value. Master data management solutions. This is the language of the future of business. As the amount of data companies attain and store grows, so too must the ability to deal effectively with this digital avalanche. Michael Goul is a professor of information systems at the W. P. Carey School of Business. Here, he discusses how businesses will have to learn to manage unprecedented amounts of data as a means of gaining a competitive edge.

Transcript:

Knowledge: Digital universe. Exabytes. Data fluidity value. Master data management solutions. This is the language of the future of business. As the amount of data companies attain and stores grows, so too must the ability to deal effectively with this digital avalanche. Michael Goul is a professor of information systems at the W. P. Carey School of Business.

Here, he discusses how businesses will have to learn to manage unprecedented amounts of data as a means of gaining a competitive edge. I recently heard a new phrase related to how the world is creating more data than ever before. Can you explain what is meant by the phrase "digital universe?"

Michael Goul: Sure. There was a recent IDC study that coined that phrase by examining worldwide growth and the amount of data that's being created and stored by individuals and governments and organizations. And they're estimating that this digital universe that we're currently saving and storing is at about 218 exabytes. [laughs] That's kind of a new term. And that's equivalent to 261 billion gigabytes. They also estimate that, by 2011, the magnitude of the digital universe is going to grow 10 times its size it was in 2006.

So that's just five years for a tenfold increase. You could imagine, if we predicted a tenfold increase in the number of cars on the road, what that would mean to our transportation systems — it'd be a nightmare. Well, the same potential overload is causing organizations to have to ramp up new data-management approaches. Things like digital video, RFID, digital snapshots, and more and more customer and supplier data capture are some of the main instigators for this growth.

Knowledge: So, what does this mean to business organizations?

Goul: Well, what it means is that new digital management solutions have to be the result of a more centralized strategic planning process in most companies. But whatever strategy an organization chooses, it really has to take into account what I've been calling "data fluidity value." I know that's a complicated term. In other words, it has to recognize that data has little value unless it gets to the right people at the right time and in a context they can understand.

Most significant here is this concept or notion of fluidity. That means that the data is pliable, it can be readily reshaped, and it's mobile. That notion extends to systems, too. In other words, automated data flows between an organization's different information systems have to be such that there are appropriate translations, and so those different systems can talk and use the data.

It used to be that integrating those different systems could be done with tools that linked each one, one at a time, but that just didn't scale. It's really no longer the case today, because you can't predict who your business partner might be tomorrow. So you have to learn to be able to scale on-the-fly, one-by-one hookups, and you have to do that all at a breakneck pace in order to compete.

Knowledge: Can you provide an example of that?

Goul: Sure. If you think about Lowe's, that's become the place to go when you want to tackle a home remodeling project. So, suppose we're going to redesign our kitchen. Wouldn't it be great if we go into Lowe's, we see a simulated layout of our existing kitchen space, with a particular cabinet type, different counter-top choices, different brands of mounted stove tops, refrigerators, sinks, and dishwashers — the whole gamut.

Then, with the click of a mouse, and drag and drop capability, we should be able to see how a different cook top might look with a particular-color granite counter top. Similarly, we might want to move a cabinet and see if the refrigerator we'd like would fit. While that would streamline our design and enhance our customer experience, what's underneath and the systems to support that scenario are really complicated.

Our suppliers have to provide images of their product using a scale that fits with the scale the other suppliers are using so that oven would go in the slot underneath the appropriate counter top. If we change our product, the update has to work with the simulated layout, and it has to happen right when we stop manufacturing one product and introduce the next. And that has to be sent out to all Lowe's locations at the moment that we make that change.

But of course, as a supplier, we're probably also selling our products through other outlets as well. So that means we have to streamline being able to get that product data ready and out to every business partner that we have whenever we make that change. If we don't have the capability to do that, our products aren't going to be included in the customer experience at a company like Lowe's. We'll just miss out. So that's what I mean by data fluidity. Data has to be packaged and delivered in a way that brings value to the organization.

Knowledge: What's the ROI of managing data well? In other words, how are organizations putting their arms around your concept of data fluidity value?

Goul: Well, value lies in the use of data to drive business. So, for example, value could come from using data to support growth. I'm thinking of Andy Grove now, a former COO of Intel, who recently wrote about how large organizations need to come up with some new growth strategy because many of them are stagnating. And to do that, they can use their competitive capabilities to enter new growth markets. I think I saw in the Wall Street Journal today that GE is going to shut off their appliance area and move into the more profitable areas.

One of the things that you have to have, though, is basic competitive capabilities to enter those new growth markets. Grove suggested, for example, that GE begin shedding some of its lower-margin businesses, like I said — and they did that, [as reported in] the Wall Street Journal. And he cited Apple's move to the music industry as an example of what other companies need to do.

That means that the data a company is managing today might be different than [the data] it manages tomorrow. And that capability has to support getting value from data, whenever and wherever it's possible, in the context of the business decisions that drive that value can be made or supported. An example would be, on a mobile phone, being able to pull up the brand-new product to show to the customer, while you might be a plumber being fixing a sink at that location.

Being able to scale that and do that quickly, that type of cross-selling or up-selling could be a real opportunity. Most companies today are going after this data fluidity, investing in a suite of products that's called "master data management solutions." I'll just call that MDM for short. Those solutions involve engines that are designed to coordinate the flows of data between different systems and people.

Now, while that doesn't seem too conceptually difficult, they're amazingly complex in what they have to be able to do, how fast they have to do it, and how much data volume they have to be able to handle. For example, you might have a whole bunch of rules about how data gets translated for one supplier — say, Lowe's — and then for a Home Depot or other types of entities. And all of the data structuring changes that take place — let's say, you add a knob or change the description of a knob — you have to do that, and maintain the consistency in all that data.

Getting a solution to work is really a difficult and arduous task. And that really gets compounded if the data in your organization isn't too clean to start with. So, thinking about that Lowe's scenario, if you use a point and a click to add or delete a different dishwasher, that really represents a sale opportunity. If you're a dishwasher supplier and your product isn't in the system, there's just no way to make a sale.

Knowledge: So, how are organizations putting MDM solutions to work today?

Goul: Well, from my understanding, I've found that most companies are establishing a beachhead when it comes to MDM. For example, they may first focus on only one of a series of generic "value streams." So, really, a value stream, to use that term, is a sequence of activities and information flows associated with whatever an organization's business model might happen to be.

But there are some generic value streams that I've seen MDM in action today. One is what I call "prospect to customer." That one involves, if you have a potential customer, they're a prospect first, and how you convert them into a customer is a value stream.

This is consistent with the Lowe's example, where the organization gets value from delivering the right data at the right time in a customer experience in such a way that that person that's using that system becomes a customer and makes an order. There are other types of value streams, and some involve the supply chain. For example, a value stream of "awareness to prevention" could involve how well a company's systems are linked together with the supplier's partner systems.

So, for example, if someone is having trouble producing a particular widget that goes into our product, if they could notify us in advance that they're having that problem we can plan for it, and detect and communicate the problems and how to solve those problems, and that will reduce the risk and the cost to everyone involved in that supply chain. So that's that the "awareness to prevention" value stream. Also, when a company brings on board a new supplier, there is what's called a "relationship to partnership" value stream that becomes significant.

That one is relevant to the service industry as well. I have my American Express Costco card and American Express Dillard's card, so when American Express partnered with those different organizations to create a card uniquely for them, I am sure they relied on an MDM solution to reconcile their communications, and streamline all of the customer charges, and make sure that their partnership contractual agreements were arranged for.

Knowledge: Why are companies investing in MDM in a piecemeal fashion, for example, one value stream at a time?

Goul: That's a good question, and in fact it's kind of perplexed me. You'd think that companies would just go into this completely and say this is something we have to do, and invest in it and get it all done at once. But I think there's a combination of several factors why they're going one value stream at a time. First, IT investments nowadays require really a proof of concept. So, you create a pilot study and if that pilot study is successful, then you give it a green light to be scaled to the enterprise level.

Second, MDM is really complex, and there are a great deal of lessons and best practices that have to be learned along the way. So, for example, if you're trying to build a capability to do MDM, it's not going to happen over time. You probably have to be out there working on it to build some of those capabilities and those learnings. You probably need some MDM champions, and those would be the people with the insights and the skill sets that can really communicate and deliver a vision for the entire enterprise.

[These champions] have to come from having [implemented] it in some specific area of a value stream. So, you pick an important value stream, show its potential, show that it has ROI, and then you begin to consider how well it will scale to the enterprise level. Also, the vendor landscape is changing really rapidly. Data warehouse vendors have merged with large entities like IBM, Oracle, and SAP. We can think of Cognos, Business Objects, Hyperion — all those names have pretty much disappeared because they've merged with these larger companies.

And others are ramping up to serve the small to medium enterprise market for MDM. What that really means is that organizations are taking the attitude that if we can get in with a vendor, work closely and build a solid relationship with that vendor and that solution provider, then that would be a good step along the way of locking into a move to the enterprise level solution.

Knowledge: Do you foresee any problems or issues with this piecemeal approach?

Goul: Yes. And an analogy might help to explain why — if you bear with me for just a minute. I'm going to go back to the Ice Age here. Back in the Ice Age we know that there were glaciers all over the place. And when the earth would heat up underneath those glaciers through volcanic activity, a pool of water would form underneath the glacier and then actually cause it to float. And once it was floating, then the natural dam that that ice was providing would be lifted just enough to allow a huge amount of water to gush out and cause a huge flood.

In fact, I come from the Pacific Northwest and that's how some of the area in Eastern Washington was created, from these types of floods. And in fact, the debris and sediment gets thrashed about as this tsunami-like wave comes down from those glaciers into the lower elevations. There's an Icelandic word for this, and I'm sure I'm going to pronounce it wrong because no one, unless they speak Icelandic, can probably say it. But it's called "jokulhaups" — that's what they refer to these phenomena as.

Now, the way that this relates to MDM solution is because if you get an effective solution within a particular value stream you can pretty much be lifting that value stream just enough [to release] a flood of needs for MDM across other, integrated, and closely related value streams, that are going to need to be dealt with quickly.

To give you an example, if we go back to Lowe's, let's suppose the supplier successfully puts together MDM and streamlines the display of a product in that layout that's going to appear in the customer experience when we go into the Lowe's store.

The processes associated with that new product design, all the things that might make that happen at the company, comprise a couple of other value streams. One's called "concept to development." Someone gets a gleam in their eye and says this is what the new dishwasher should look like, and then they develop the diagrams and those types of things. That becomes a very important value stream, because in order to get it quickly to the Lowe's system you have to be able to streamline how quickly your designers can get the thing into manufacturing.

And in fact, then once an order is taken you need to make sure that you can get the distribution down right. That's the "manufacturing to distribution" value stream. If you can't deliver the product that's ordered through that customer experience you're going to lose out on the next sale.

So, if you really can't integrate these investments across your value streams in MDM very quickly and build that into a full-blown MDM capability, then you're going to be competing in a tough area, because there's going to be another company out there that's going to be faster at leveraging their first investment, and parlaying it throughout all their tightly coupled value streams more quickly than you.

Now, I've proposed this possibility to people working to deploy one value stream at a time today, and they say they like my story about the Jokulhaups scenario where the flood wreaks havoc, but they say that it's probably not as catastrophic if you develop some type of a stepwise value stream by value stream strategy that will facilitate a real smooth transition to your enterprise-wide master data management.

But you really have to think about that beforehand, because — we'll use the analogy again — once that iceberg lifts and the flood starts, if you're not prepared for it and haven't thought through it, have your champions in place, and your trained people, and all your lessons learned, ready to be articulated throughout the organization, it's going to be a slow-moving process and a competitor is going to probably beat you.

But all the people that I've dealt with and talked to out there that are working on MDM solutions really agree that this digital universe is compounding in growth at an astounding rate, and this concept of data fluidity value as a focus is really becoming a basic competitive necessity.

Managing IT Faculty