Analyst Lyndsay Wise recently published a useful BI demand-side analysis, Redefining the Mid-Market and Its Business Intelligence Requirements. Her thesis is that "small and mid-sized organizations seem to get the short end of the stick when it comes to their software needs" based on IT infrastructure she sees as needed to support BI and related technology. Yet I wonder if segmenting BI-user enterprises by gross revenue is the best way to look at the BI demand-side market. Let's consider other, more refined approaches.First alternative: If the concern is IT infrastructure — I won't get into what infrastructure is actually needed to support BI — shouldn't we segment BI-user enterprises by IT spending? For instance, a grocery chain might use BI to optimize marketing, sales, logistics, staffing, store siting, and so on, but ultimately that chain is in the business of selling groceries. An on-line focused retailer will spend proportionately less on physical facilities and more on IT. Each might be an SME (small/medium enterprise) based on revenue, but their capabilities to support comprehensive or spot BI are very different.
What of large, decentralized enterprises, where IT is managed on a component level? Take an extreme example. The U.S. government's budget is over $3 trillion yet the overall budget of the Bureau of Labor Statistics, which like most U.S. government agencies largely operates its own analytics infrastructure, is around $600 million, placing it in the mid-market. Cases like this one, viewed for the purpose of BI demand-side market segmentation, can be found throughout industry and government. They suggest that looking at the independently managed components of larger enterprises, classifying those sub-entities by size if desired, would provide a more useful segmentation than treating many large entities as if they were monolithic.
Another angle concerns what I'll call analytical styles. Again consider an example, going back to my grocery chain and my on-line focused retailer. The bricks-and-mortar enterprise will look to traditional BI interfaces for its analytical needs. The on-line enterprise may look to those same applications in managing internal operations, but it's also likely to wish to build BI capabilities into its customer-facing interfaces. The BI tools marketplace and the marketplace for end-user BI applications are very different. On the tools side, there are a plethora of powerful, low-cost options that will allow you to program BI into a line-of-business (LOB) application. You don't need a comprehensive BI applications/solutions platform to fill this LOB need effectively and (relatively) cheaply.
Slicing up the BI demand-side (user) market by revenues is a start. I've suggested three approaches that can lead to a more refined picture. We can further segment the market by organizations' actual IT capabilities and spending. We can look at organizational components that operate independently. And we can consider the type of BI required. These refinements are a continuation. It's important to understand the BI market. Start with assumptions that make the analysis tractable, but do also acknowledge that business models vary and that accurate, useful analyses may be complicated. In the end, an analysis that is over-simplified may prove shallow.Analyst Lyndsay Wise recently published a useful BI demand-side analysis. Her thesis is that "small and mid-sized organizations seem to get the short end of the stick when it comes to their software needs" based on IT infrastructure she sees as needed to support BI and related technology. Yet I wonder if segmenting BI-user enterprises by gross revenue is the best way to look at the BI demand-side market. Let's consider other, more refined approaches.