Enterprises – and vendors – are still trying to figure out where mashups make the most sense.

The most popular, low-hanging fruit for mashups seems to be replacing portals, or at least making them more Web 2.0-like, and smaller data-integration tasks that don't quite rate as a priority, but nonetheless need to be done for some business purpose.

One use that seems to be capturing the enterprise's collective imagination is the idea of using mashups for business intelligence applications, according to a recent TechTarget article, "Enterprise mashups: Tools build data integrations."

As the article points out, one of the good things about mashups is you control how far you want to go. When it comes to using mashups for business intelligence, Forrester Research analyst James Kobielus says there's four types of BI mashups, starting with lightweight, presentation mashup used to  generate reports and ranging to a still-theoretical collaborative mashup where data is stored centrally and governed by IT policies.

Kobielus says mashups aren't something IT can force on users. The drive for adoption needs to come from users, and that means making it fun, highly interactive and – of course – easy.

Kobielus co-authored a white paper on the topic, "Mighty Mashups: Do-It-Yourself Business Intelligence For The New Economy," which I found  for free on Inetsoft.com's site. The report makes it clear business users are hungry for the sort of self-service, down-and-dirty data mashups can provide:

“Though still immature, BI mashup is well on the way to becoming an enterprise-ready approach for self-service development. The key enablers for enterprise-grade BI mashup are IT governance practices such as design templates, version controls, and monitoring. When supported by strong governance, information workers can mash up personalized views from reusable analytic components that are developed and managed for them by IT. Mashup enables nontechnical users to build dynamic views of disparate data that are personalized, context-rich, role-tailored, and ad hoc to explore this data in greater depth.”

You would think this would make mashups a shoe-in, but as with all things techie, politics and corporate culture can gum up the works. In part two of the TechTarget enterprise mashup series, Pfizer Research Fellow Michael Linhares relates how culture attitudes had to change first. Because of the importance of research findings, there were those who didn't want to share their data. What's more, IT had to expand its mindset to realize that not all infrastructure tools are IT tools, the article explains.

But if you can overcome these political and cultural issues, data mining can be a great way to justify your investment in mashup tools, according to Michael Ogrinz, an enterprise architect and author of "Mashup Patterns: Designs and Examples for the Modern Enterprise." In a podcast, also published by TechTarget, Ogrinz says these types of mashup projects can deliver a quick ROI. In fact, he felt strongly enough about it that a good deal of his book focuses on data mining mashup patterns.

The TechTarget article lists a number of companies offering enterprise mashup products. Kobielus recommends those interested in enterprise BI mashups check Microsoft Project Gemini and IBM Cognos.

In my experience, analysts are a bit shy about recommending specific vendors when talking to journalist, so the fact he mentioned specific vendors and prodcuts caught my eye. His reason is that those are the two products he felt were far enough along to allow companies to reach that level four mashup maturity, where you can build collaborative mashup, data is centrally stored and the mashup permissions are governed by IT policies.