Anyone who refers to their key audience as coneheads deserves to wear the dunce cap.

When he launched Microsoft's Popfly mashup tool at a Web 2.0 event this week, Microsoft CEO Steve Ballmer used "conehead" to describe the traditional programmer. He didn't bother coming up with a word for the people at whom Popfly is actually being aimed. That's because, for all the make-your-own-app kits that enterprise companies are creating, no one really has a common profile for this kind of customer. It's someone Web-savvy enough to understand widgets but not experienced at database integration. It's somebody who has an expertise in productivity software but who couldn't be trusted to handle an Exchange migration. People who are sophisticated enough to know what they want from technology but not stupid enough to assume they can get it without some IT department involvement.

As promising as Popfly and the other mashup toolsets are, Microsoft and similar vendors haven't appeared to have done a lot of thinking about their go-to-market strategy. Just offering these kinds of products is not enough. As we've learned from experience through traditional desktop software, a lot of the most interesting features never get discovered, let alone used. Similarly, mashups will require a different kind of marketing, and the most logical channel is through the IT professionals who can explain the opportunities they open up and the limits to their functionality. And there are limits, although Microsoft and Google don't say that too loudly so as not to rain on their own parade.

Popfly can allow widgets and or min-apps that tie into traditional Windows programs like Word or social networking platforms like Facebook. That means they are most likely applications that do simple information publication and exchange, stuff that doesn't take up a lot of compute resources. They might pull from enterprise systems, but IT managers would be negligent if they allowed them mashups free rein to change the way such systems operate. The most useful applications will still probably be delivered by experienced programmers well versed in handling large data sets, languages and topologies. There are plenty of people outside the finance department using Excel spreadsheets, but that doesn't put them on a par with CFOs. Similarly, mashups will foster primarily second-tier applications that focus on individual needs but not necessarily advance the objectives of the business as a whole.

Microsoft and other vendors could take some time out from making potshots at one another and show some real leadership by helping IT managers understand how and when they should train users on mashups. In some cases it could be a way to increase their own productivity. With other business processes and systems, you need more oversight. The nature of "acceptable use"in terms of applications may undergo a major change. By offering the tools without that kind of education, the cart is being programmed without the horse.

No one is really suggesting that mashups will do anyway with the IT department, but they still manage to ignore it, as though non-programming users would be appropriate by acting in isolation. While he doles out nifty new toys with which the average enterprise employee can build applications, people like Steve Ballmer would do well to pay a bit more attention to the coneheads.