Data services will shift from something you primarily see in large organizations to wider deployment in small- and mid-sized companies, according to a recent set of predictions by Forrester. One reason: Using a data services approach for integration can bring in data from more sources while reaping a return on investment in a short time, says Forrester analyst Noel Yuhanna.
If you're not familiar with the term "data services," you may find it helpful to think of it as a way to liberate your data from specific applications. It turns your data into services, which can then be called upon by different applications, portals, even mobile devices. It's actually a natural extension of service-oriented architecture, which did the same thing for application capabilities.
You should know that Forrester prefers the term "information as a service," but it's the same thing as data services. I happen not to like that term, because the acronym-IaaS-can be (and has been) confused with "integration as a service."
If you'd like to see just how complicated, Yuhanna recently participated in an excellent webinar explaining data services, aka information-as-a-service. In the Denodo-sponsored webinar, he suggests you think of data services not as one approach or one solution, but rather as a platform, supporting agile data integration. He even has a nice slide that shows you the components of the platform. You'll find it at the 15 minute mark (you can slide to 15 if you download the slide file). He also has a great slide that shows you how different solutions-EAI, ETL, BPM, Content, IVR-feed into the platform. It virtualizes, and thus hides, all that data integration complexity.
But if you're more interested in the business value, here's what he says is the take-away: Data services simplifies data integration, allowing you to integrate information from more sources in less time. And it makes your data more widely available-even to the point of self-service.
I know, it sounds like a lot of hype. And, like I said, it's complicated underneath the hood. And while a completely integrated data services platform could take years to do, it doesn't have to be this massive, boil-the-ocean undertaking, according to Yuhanna:
You are actually liberating information from the stovepipes. ... The good thing about this is the data could be coming from a few sources or many sources, and you build this virtualized framework to deliver different types of data. This is what we call data services. The good thing about this is the quickness, the ROI, the benefits of doing this very quickly-even for smaller SMBs companies who need to have integrated data quickly now to use for different applications.
He offers several examples of how specific businesses have used data services to address problems, including:
But it's the ROI and quick turnaround that make this approach appealing to organizations of all sizes.
"You don't have to spend million of dollars doing this. It doesn't have to be a six month project. You can do it in days," Yuhanna says in the webinar. "The idea is you want to decouple application from the data. Why should the applications be very tightly coupled, right?"
If you'd like to read about how others have gotten the ball rolling on data services, check out this information management article by Robert Eve of Composite Software. Eve explains how three companies-an energy firm, a bank and a telecom-deployed data services, and in one case, in place of a data warehouse. In all three cases, they established an Integration Competency Center to help them manage the process.