Not too long ago, I shared a list of business intelligence best practices from Gartner. The analyst firm produces consistently excellent research on BI, which is why an InfoWorld article on its BI predictions for 2009 caught my attention. While the predictions don't involve anything really new, there are some interesting insights nonetheless.

First, companies will look for more direct business value from their BI initiatives. Thanks to the tough economy, this is a trend that will apply to pretty much any IT investment, including CRM, as I wrote last month. But companies won't find it easy. Through 2012, says Gartner, a quarter of the largest 5,000 companies will regularly fail to make insightful decisions about important changes in their business and markets.

Gartner also predicts business users will assume more responsibility for BI, controlling at least 40 percent of the BI budget by 2012. Excluding IT and purchasing packaged analytics rather than platform solutions could result in siloed applications and data, "which will limit cross-function analysis, add complexity, and delay corporate planning and execution of changes,"says Gartner analyst Nigel Rayner.

Ted Cuzsillo echoes this prediction in an Enterprise Systems article, noting that it will result in increased adoption of new and highly individualized tools from companies like Panorama and Tableau. "Concerns about self-service chaos will be ignored for now," he writes. Not only that, but companies may cut their analyst ranks in favor of recruiting more business users. Cuzzillo mentions BI will "look more like an everyday business function."

Three other predictions from Gartner:

• Twenty percent of companies will use a vertical SaaS-style analytic application by 2012. Some companies use these types of apps to supplement existing BI investments, as I wrote in March.

• Companies will experiment with adding social software to their BI mix. I recently wrote about how Google and other companies are doing so with internal prediction markets.

• Companies will increasingly deliver analytic applications via coarse-grained mashups, a simpler and more flexible approach than building composite applications out of fine-grained services.