Technology vendors like to talk about platforms, because platforms imply a broader footprint both in terms of functional capabilities and in terms of implementation usage. Platforms also sound more “strategic,” even if the practical implications are vague. But the term “platform” can also be simple marketing hype. How do you know when a software “platform” is really a platform? More specifically, do data integration platforms exist now?
In simple terms, platforms are a foundational infrastructure or framework upon which you build things like databases or custom applications. Platforms typically develop from a heritage of individual productivity tools, coming together over time to provide a coherent set of capabilities targeted at addressing a class of problems. Platforms have impact when used across wide swaths of the organization, increasing productivity and reliability by fostering standardization and reuse. Several types of technology platforms are well-established and proven. Database platforms are very mature, and application platforms, particularly Java-based application development platforms, solidified in the last decade.
So what about data integration? Several years ago, “data integration” as a discrete category or discipline didn’t exist. There were different technologies and tools, such as ETL, for integrating data. But data integration has been maturing, with technologies’ capabilities broadening and converging, and best practices and methodologies deepening. Moreover, the need for a common approach to data integration has become increasingly apparent in recent years. Quoting Gartner’s recent report, Predicted 2009: Technology Changes Will Shape the Future of Data Management and Integration: “Organizations increasingly see the uncoordinated and reactive approaches they have used over the years as impediments to progress in addressing business pressures and the widespread information-glut that negatively impacts operational efficiency. The lack of integrated tools inhibits organizations’ ability to respond efficiently to changing business requirements.”
Put simply, many organizations still take an ad hoc, project-by-project approach to data integration, with each project selecting its own tools and methodologies. And this ad hoc approach is simply too expensive and too complex and too risky. Organizations need to adopt a more standard approach to data integration—including the technologies they use and the methodologies they employ. The technology underpinnings of this standard approach is a data integration platform.
Here is Informatica’s definition of a data integration platform: “The comprehensive set of technologies for accessing, discovering, cleansing, integrating and delivering timely, trusted data to the extended enterprise.” It’s the technology infrastructure upon which data integration project types, such as data warehousing, data migration, enterprise data quality, MDM, data synchronization, cloud data integration, and B2B data exchange, are built.
Does any vendor have a perfect data integration platform? Of course not, but at Informatica, we think we’re the furthest along. In the next few postings, I’ll be discussing the criteria for a data integration platform based on the common customer requirements we have seen. I’d also be interested in hearing your thoughts about what a data integration platform is, or whether you even care.