Sybase IQ

« Summer School | Main | Sybase IQ – Making Data Easy to Consume »

July 15, 2010

Just Like TV

Over at ITBusinessEdge, Mike Vizard presents some interesting analysis of why more widespread use of BI is not occurring within organizations, and what he thinks will ultimately drive higher rates of adoption:

There are plenty of reasons, ranging from cost of application licenses to reliance on spreadsheets. But the biggest issue might be that using a BI application seems unnatural — the data first has to be imported, then users have to leave the application environment in which they are working in to access BI.

According to JasperSoft CEO Brian Gentile, application developers are rapidly becoming more aware of this issue and increasingly embedding BI directly into their applications. Open source makes it easier and less expensive to do so without having to develop that functionality themselves.  Gentile says BI capabilities are well on their way to becoming a standard component of other applications.

This doesn’t necessarily mean the end of stand-alone BI applications. But the number of people accessing BI software within another application environment will dwarf the number using stand-alone BI software, who tend to be business analysts.

There is no question that enterprise application vendors have added and (will continue to add) BI capability to their environments, from simple reporting and dashboards in some instances to fairly sophisticated analytical capability in others. And Mike may well prove prophetic in his assertion that there will one day be more people accessing BI capability via these applications than via dedicated BI software.

Interestingly, that doesn’t mean that there will be fewer people accessing dedicated BI and analytics software than exist today. Let’s analogize for a moment and liken analytics capability to video content. Like analytics, video content can be accessed in a number of ways. You can use a TV, which is a dedicated environment analogous to BI / analytics system, or you can access video as an auxiliary function of some other device — laptop, smart phone, SUV — which is similar to accessing analytics capability within an enterprise application. Although more and more people are accessing more and more video content via these other methods, the number of TVs and the number of people watching them has hardly gone down. If anything, it has increased significantly.

Likewise, the number of users of analytics systems is increasing rapidly. Some of these users may, indeed, find it “unnatural” to have to leave the application in order to do analyses on their data. On the other hand, most of them were already leaving that environment to access those incomplete, inconsistent, and time-consuming spreadsheets. The reality is that it isn’t just the number of analytics users that’s growing, it’s the level of sophistication that all BI and analytics users require. As these users grow more sophisticated in their requirements, many will make analytical demands that enterprise systems will have a very difficult time fulfilling.

Why? Well, first there’s a problem of scope / feature creep. Development teams supporting enterprise applications find it challenging enough just keeping up with the demands placed on the core business processes they support. They are happy to provide reporting and analysis of data related to these processes (up to a point), but in light of the demand for increasingly broad and deep analytics capability, left unchecked they might ultimately find themselves supporting analytics / BI systems which also happen to have some operational features!

Moreover, as analytical needs grow more sophsiticated, the importance of having the correct underlying infrastructure becomes increasingly clear. One day we may well have a data management technology that can be simultaneously configured to provide optimal results for both analysis and transactions, but we aren’t there yet. Today, putting both functions against a single data source requires making sacrifices either to transactional performance, analytical performance, or both. With demands for both analytical and transactional performance growing rapidly, users are not likely to embrace those sacrifices.

So, yes, BI and analytical functions will continue to be embedded within enterprise applications, and the number of people using these capabilities will grow rapidly. At the same time, dedicated BI and analytical systems will  be more widely accessed than ever before, supporting increasingly robust and sophisticated capabilities. One does not come at the expense of the other, any more than millions of Youtube views cut into sales of widescreen LCD HD TVs. They both just keep growing together.

Posted by Sybase IQ at July 15, 2010 3:45 PM