Nothing but .Net? Nyet!

by Geoff Hart

Previously published as: Hart, G. 2001. Nothing but .Net? Nyet! http://www.techwr-l.com/techwhirl/magazine/usersadvocate/usersadvocate_net.html

For the past year or so, Microsoft has been heavily promoting its so-called “.Net” strategy. Among other goals, this strategy aims to shift at least some of Microsoft’s enormous audience to a model in which they rent Web-based software rather than, as in conventional marketing, purchasing a CD and annual upgrades. The claimed advantages of this strategy for users include the fact that we’d pay for software only when we used it, and that we’d always have the most recent release of the software available, including all the latest bug fixes and security updates. Microsoft’s Steve Ballmer has publicly stated that in 10 years, software rental will have largely replaced purchases. Other aspects of the .Net strategy revolve around Internet-enabling software such as Microsoft Office to work with other Web software, which is also a laudable goal.

But it’s the notion of renting software over the net that interests me most, since it has two important lessons to teach us. The first of these lessons is fairly obvious: Microsoft is trying to join the growing group of entrepreneurs interested in earning their dotcom millions by serving as “application service providers” (ASPs). The problem with ASPs? They make sound economic sense, but ignore the needs of a majority of the software’s community of users. That’s hardly wise, and the problem lies in something that any reasonably skillful technical communicator could identify with a little thought about the needs of their audience.

Why would this approach fail? After all, given Microsoft’s track record, it would seem awfully foolish for me to bet against them and those who will follow their lead, and the idea does seem superficially reasonable—but despite this, I predict that the ASP aspects of .Net won’t work nearly so well as Microsoft hopes and may even fail outright. The problem is that their strategy is driven more strongly by economics and a fear of competition from smaller, more nimble ASPs than by customer needs.

The logic behind Microsoft’s ASP approach is that an annual rental or a “per use” fee guarantees the company more income than with the current model. After all, you can bet that Microsoft isn’t adopting this approach to lose money. But even if rental fees seem to cost average users less than annual upgrades, many of us won’t see these savings because we don’t upgrade our software annually. The reluctance to upgrade stems from a very real fear of encountering new bugs once we’ve finally learned to live with the current batch, and a prudent “if it ain’t broke, don’t fix it” philosophy. With an ASP approach, we’re required to pay annually even if we’d prefer to use a single version of our software for several years.

Which other needs does .Net ignore?

First, many of us will refuse to embrace a solution that forces us to remain connected to the Web at all times to run Office 20xx, particularly before broadband access replaces dial-up connections. Laptop users will rebel against being tied to a power socket and an expensive wireless connection—something that would prevent computing in places such as subways, airplanes, or cabins in the woods. These obstacles will gradually diminish, but won’t go away entirely, nor cheaply and soon.

Second, most users of Microsoft products feel that once we've finally got a stable and productive configuration, we want to keep that configuration. Imagine the horror of working with a new and “improved” version of Office every time you log on! The ongoing bug fixes and security updates will be nice, but Microsoft’s "improvements" have often caused more problems than they solved; their abominable record of service releases for Office leaves little room for optimism.

Third, in the wake of Melissa and the “Love Bug” virus, no MIS manager and not many software users will cheerfully accept an approach that relies on Microsoft for network security. (Widespread concern over Microsoft’s “Passport” product arises from these same problems.) Sure, firewalls will provide additional security, but the number of back doors this approach would open into corporate networks is downright scary. Until Microsoft demonstrates that it takes security seriously, adoption of .Net will require a leap of faith many managers won’t be prepared to take.

So for these and other reasons, .Net won’t replace traditional software distribution any time soon. The strategy will consume considerable resources better spent finding out what customers really want and providing it. We’ll soon see whether Microsoft can ignore user needs this badly in pursuit of their own vision, unlike those of us who write for a living and understand that we can’t ignore our audience.

And therein lies the second, less obvious lesson for technical communicators. Microsoft makes lots of noise about how well they listen to the users of their software—yet they still haven’t fixed longstanding, well-known, loudly demonized bugs such as the problems with autonumbering in Word. Microsoft is justifiably proud of the usability testing labs they’ve created, which have indeed helped them produce software that greatly empowers us, no matter how much we love to complain about its failings—yet despite this focus on usability, they don’t seem to listen to the much larger community of users outside those usability labs. In both cases, development managers at Microsoft have consistently failed to find out what their audience really wants—or failed to listen to that audience if they did hear its many voices.

One thing that scares many people, including Microsoft (one assumes), away from attempting an audience analysis is the perceived complexity and cost of such an exercise. Yet in the modern Internet age, with discussion groups available on just about any topic you care to name, it’s trivially easy to learn what your customers are saying about you and your product: all you need to do is find and monitor discussion groups that touch on your product. For example, a quick look at the archives for techwr–l, Word-PC, and Word-Mac, among others, would reveal a wealth of real-world data on the problems users have with Word, the types of questions they repeatedly ask, and the ways in which they try to accomplish certain tasks. This is the kind of information that formerly would have cost many thousands of dollars to collect, but which is now available at a cost of nothing much more than an employee’s time to monitor the discussions and mine the archives.

This second lesson is important. While I’m deeply skeptical about certain aspects of the “dot Net” strategy, there’s enormous value in a strategy that uses the Internet as a source of audience information. That kind of a “net” strategy is something you don’t have to be as big as Microsoft to take advantage of, and something any technical communicator could benefit from.