A growing number of companies are saying goodbye to "stovepiped" approaches to integration and looking to redesign data integration strategies with enterprise-wide goals in mind, according to Cambridge, Mass.-based IT analyst firm Forrester Research Inc.
The growing movement toward more "cohesive" or "holistic" integration strategies is largely the result of major changes in the world of data management, such as the growing popularity of things like mobile computing, cloud computing, data virtualization and "big data" analytics. It also has a lot to do with the desire among organizations to gain a higher level of control over both structured and unstructured data, said Mike Gilpin, a Forrester vice president and research director.
Gilpin -- who based his statements on the results of several recent Forrester surveys -- said organizations today increasingly want their business processes to be informed by the right data and they want applications to be both sources and consumers of information. Finally, he added, organizations want the ability to deliver data to a variety of applications regardless of where the information is sourced in the underlying environment.
"It used to be the case that you would take one set of existing technologies and tools and do application integration, another set to do process integration, and another set to do data integration," said Gilpin, who spoke during a recent webinar hosted by open source data integration software vendor Talend. "But all of these cross-silo approaches are driving the need for a more holistic [strategy]."
Gilpin warned, however, that with all of the changes happening in the world of data management, implementing a holistic data integration strategy that works for the entire enterprise can be difficult.
"But that doesn't mean that you shouldn't have a cohesive strategy," he said. "What's really necessary today is a strategy that is designed to cope with this level of change."
Sallie Mae seeks a more cohesive data management strategy
Wilkes-Barre, Penn.-based student loan provider Sallie Mae Inc. is just one example of an organization that is responding to changes in the realm of data management with a few changes of its own, according to Michele Koch, director of the company's data governance office.
The need to gain valuable business insights from unstructured data found in email, documents and on the Web is driving Sallie Mae to implement a more cohesive "information architecture," Koch said. The company hopes the move will, among other things, give business users the ability to more easily integrate information from various source systems.
The phrase "information architecture" refers to a well-defined system of data management that stipulates how data is to be stored or grouped while making it easier for users to find the information they need. The overall design of the information architecture will vary depending on the needs of individual user organizations, according to experts.
Koch agreed that changes in the realm of data management can make implementing a cohesive strategy difficult. She added that the sluggish economy poses another challenge.
"We were so focused on structured data all these years, like databases, and now people are looking at unstructured data," Koch said. "It's just very hard to do because we have less staff and resources than we ever did before."
Tips for creating a more cohesive data integration strategy
Organizations that want to create more holistic data integration requirements should begin by identifying and focusing on areas in which a cohesive data integration strategy will provide the most value. This means looking at where the most significant investments are happening and gaining an understanding of what kinds of integration are necessary, Gilpin said.
For example, many companies today are looking to synchronize information between mobile devices and back end systems, thus making it easier for employees to conduct business on the road.
"Another common pattern might [involve] working with a trading partner with whom you need to interchange business documents that represent transactions formatted in a standard industry representation," Gilpin said.
After identifying key integration patterns, it's time to look closely at the various integration methods and come up with a standard set of recommendations for tackling each project. The idea is not to come up with a mandate, but rather a series of integration choices that are "modular in nature" and can plug into one another.
There are several benefits to having a standard set of choices or recommendations for each type of project, Gilpin said. For starters, it serves to reduce the number of technologies in an organization's application portfolio, thus making the portfolio easier to manage. A standardized approach also helps developers move more quickly on integration projects by providing them with a "jumpstart." Finally, Gilpin said, standardized approaches will ultimately mean lower costs associated with training employees and acquiring people with the right skill sets.
The next step is to create an integration competency center, Gilpin said. The group should include representatives from both the business and developer community who work together in an effort to maximize the value of the organization's recommended integration tools and methodologies.
"In organizations that are more distributed in nature, where the competency center model is not the favorite approach, we also see integration builds being developed and managed within integration communities of practice, which are more loosely coupled," the analyst said.