“Enterprise analytics” is a widely used term these days. As often happens, though—it’s being used in different ways, by different groups, for different reasons. Enterprise analytics can refer to any or all of these three concepts:
1. Access to analytics capability (so users throughout the enterprise can perform their own local analytics)
2. Access to enterprise-level analytics (so some users can see reports or dashboards that incorporate data from the whole enterprise)
3. Analytics platforms that can function at an enterprise level (working with multiple data sources and formats)
Consultants, business writers, software companies, and IT execs may all be using the term enterprise analytics to meet their own communication needs—so conversations can get a little complicated, and research can be somewhat confusing.
B: Why does it matter?
In each of the three ways listed, enterprise analytics can provide an important solution to a serious problem. But each problem is different. Version 1 solves the problem of users who can’t get the kind of business insight they need because they don’t have the right tools, and may not have access to the data they want to analyze. This type of enterprise analytics speeds up the analytics process and makes it more relevant to real business issues.
Version 2 can help with the problem of data silos, in which data for different departments, divisions, product lines, etc., is stored and managed differently. Although the silo problem has diminished in recent years, it’s certainly not past history. (In a recent Deloitte poll of 1900 technology executives and business professionals, more than half of respondents cited “departmentally siloed information” and “limited cross functional interaction” as the primary reasons for inadequate business intelligence.) Enterprise analytics can be designed to overcome or at least compensate for these data disconnects.
Version 3 addresses the problem of disparate data at an even more fundamental level. There may be multiple databases and/or data marts scattered through a company. And important information may be kept in spreadsheets or vendor databases or legacy systems—all cut off from interaction with enterprise-level databases. An enterprise analytics platform can (to some extent, at least) utilize data from a wide variety of repositories.
Any of these solutions can make a huge difference to an organization. All of them together can be transformative, if they are properly integrated and implemented. But it’s important to keep in mind that the word “enterprise” is functionally equivalent to the word “huge,” so developing and delivering enterprise analytics is a very big project, no matter what the definition.
C: What’s next?
The enterprise analytics challenge is growing, because companies must deal with steadily escalating volumes of data. Not only are organizations collecting and storing vast quantities of internal data, most must now add web data to the stream of business information. At the simpler end of the web analytics spectrum, data is gathered by page tagging and/or log files, then analyzed after the fact. Increasingly, however, there is a need to perform real-time analytics, which means adding more sophisticated processes such as business activity monitoring (BAM) and complex event processing (CEP).
To learn about Spotfire’s intuitive enterprise analytics solution, check out Gil Allouche’s recent webcast “Introduction to Spotfire Analytics”.
