In a recent article, CIO.com said that analytics and BI will be the top technology priorities for CIOs in 2012, based on a Gartner Inc. survey of IT executives. However, if you look back in time, reports show that BI was a top priority even then. Although we have fast-forwarded many years, the priorities haven't really changed. BI is still top of mind.

Granted, the amount of data that needs to be processed is growing by the day, and the need for businesses to have timely insight into things that matter is becoming more immediate. But wasn't this the case earlier as well? Businesses have always had this mindset - hence the reason for growth and continuous innovation.

What's new? Nothing, on the face of it. Except that with all things being equal, the fundamental problem, or shall I say problems, seem to have taken a backseat, yet again. We seem to keep talking about the symptoms instead of treating the issue at hand. In a recent report by Gleanster, LLC, the biggest challenges for enabling BI agility, are:

• Breaking down data / departmental silos
• Integrating with applications (e.g., CRM), operations and other platforms
• Achieving acceptable data quality

The report also points out that the most commonly used metrics by businesses are time-to-decision or time-to-response to information requests; information access (comprehensiveness, accuracy, and consistency); and volume and quality or actionable insights. These, in essence, are the fundamental requirements that need to be fulfilled to the hilt in order to enable BI agility.

For those in the know, this is not something BI tools can address on their own. A recent blog by Forrester Research, Inc., states that traditional BI approaches often fall short because BI hasn't fully empowered information workers, who still largely depend on IT, and because BI platforms, tools, and applications aren't agile enough. Now that we have this background in place, I can start my analysis.

Based on what we are seeing in some ongoing polls, without the underpinnings of a self-service driven agile data integration strategy in place, BI agility will continue to remain a pipedream. Yes, of late, data virtualization has emerged as an agile data integration approach that can enable BI agility. But as all solutions are not created equal, let's try to address the challenges we discussed with the proposed solution.

As I always say, the devil's in the details. Data virtualization built on data federation does one thing and only one thing very well - it accesses and merges data from several different data sources, in real-time, without physical data movement. It can turn many data silos into one and integrate with applications. But how about data quality? Is federated data truly ready for consumption? All I hear is silence.

A BI tool won't do anything to improve data quality as it simply assumes the availability of the most current and accurate data. What happens if there are inaccuracies and inconsistencies after federating data across various systems in real time? A more fundamental question - what if you cannot effectively analyze and profile the federated data in the first place? Well, you need further processing.

Did you read the fine print? I think it just said, deal with it. Or worse yet, I have also heard the excuse - BI tools do not expect consistent and accurate data. Very convenient wouldn't you say? Bottom line, you not only lose the time advantage that you gained in not moving the data physically, but you now have to deal with quality and consistency on a reactive basis. So much for an agile data integration approach.

We discussed quality and consistency. Now, how about the role of business users? Shouldn't the analyst define business entities, analyze and identify issues with the data, create rules to correct inaccuracies and inconsistencies, and then play a key part in making sure the federated data is as requested? Ask any BI professional, business users know the data the best. Data federation does little to get them involved.

Next, let's talk about the role of IT. Is it just about prioritizing a backlog of growing requests, building out the solution, testing and then deploying it? Shouldn't IT interact with the analyst instantly and throughout the process? This is critical to IT building exactly what the business wanted. Without self-service, agility can't be ensured. However, data federation has been typically a coding-heavy IT tool.

Although data federation has been around for a long time, it hasn't gone too far. Data virtualization built on data federation seems to be a case of doing the same thing again, and expecting a different answer. Federating data across many diverse data sources, in real time, without physical data movement, is what I call, par for the course. To enable BI agility, you need to go beyond looking under the hood.

Since data virtualization built on data federation cannot profile both data sources and logic, apply complex data quality rules and advanced data transformations on federated data as it is in flight, involve the business user early and often, and reuse the virtual views not just for BI tools, portals and composite applications, but also for batch - it looks like we have a choice to make?

The choices are - manual coding, further processing using other tools, and custom solutions. Really! Is this truly a choice you have the luxury or the extra budget to make? Are you going to sign-up for a solution that promises agility and then leaves a major portion of the task to you or to another tool? What's even more dangerous is that lack of critical functionality is simply passed off as good-to-have.

The Gartner Magic Quadrant for Data Integration Tools, October 27, 2011, says it well - it's "the ability to switch seamlessly and transparently between delivery modes (bulk/batch vs. granular real-time vs. federation) with minimal rework." Data virtualization thus needs to be built-on data integration to truly enable BI agility. Having said that, I believe the days of a one-trick pony are numbered.

•   •   •

Don't forget to join me at Informatica World 2012, May 15-18 in Las Vegas, to learn the tips, tricks and best practices for using the Informatica Platform to maximize your return on big data, and get the scoop on the R&D innovations in our next release, Informatica 9.5. For more information and to register, visit www.informaticaworld.com.