What does the term “agile data integration” mean?
Personally, agile is not my favorite IT word, mostly because it looks like a word everyone knows the definition of — agile, as in light, quick, nimble, e.g., Apolo Ohno on ice — but it actually means something else.
In case you’re not from the IT side of the house, agile is a particular approach to software development. In ye olden days, an application was built in its entirety, tested and deployed — the waterfall methodology. If you wanted changes, well, you should’ve mentioned them six months ago.
Agile development is iterative and incremental, code is written in smaller batches, with testing and revisions an ongoing part of the process. So, if you didn’t see a feature this time around, it may be because it’s planned for a coming iteration of that code.
Here’s what’s interesting, though: Many companies tend to take a waterfall approach to business intelligence projects. This can make BI very, very slow, static and, ultimately, ineffective.
And who wants to wait for a new development cycle to begin when you have data to integrate now?
Aberdeen recently published research looking at how agile development can impact BI projects. The paper, “Beyond Agile Analytics: Is Agile Data Integration Next?” looked at 118 responses to a data integration survey. The top performers — which were 35 percent of those questioned — were identified as “leaders,” while the remaining 65 percent were classified as "followers."
You probably can guess what the leaders did differently. If you said “agile data integration,” give yourself a gold star! Seventy-one percent of leaders favored an agile approach to integration, compared to 51 percent for the followers, making leaders 39 percent more likely than followers to pursue incremental project delivery.
They’re also more likely to have a standard project plan in place, at 79 percent to 56 percent, and more likely to measure team performance at each stage of the process.
But here’s the thing: Leaders in this survey didn’t just “do better” than followers. They gave “followers” a serious beat-down in terms of performance. Consider:
Leaders also tend to dedicate more IT staff to integration and use a wider variety of integration technologies than followers.
The complete resource is available for free download now, if you'll provided basic registration information — although I suspect it will be relocked by the end of the week. It makes a strong case for using agile methodology for data integration and BI cases. It’s well worth the download and read if you’re curious about agile development’s benefits.