In this article from Network World, we see an interesting example of how when marketing hype meets physics, physics generally wins until someone clever enough comes along to solve the problem.

The article provides an update from the WLAN group at Siemens, who appear to have solved a pesky problem currently slowing down adoption of 802.11n deployments. It seems that the 802.3af Power over Ethernet standard doesn't supply enough power for 802.11n access points, leading to workarounds and extra costs that have made enterprise managers think twice before deploying 802.11n equipment.

802.11n has been promoted heavily by vendors and many trade media and analysts for years, even before it was standardized. The hype frenzy might lead some people to believe that 802.11n is more important to their network than electricity. This is a pattern that is repeated over and over in our industry, and even though we're all familiar with it, it's worthwhile to take a look at it now and then.

Why do we buy into the hype? From the demand side, there's an argument that technological innovation is essential to competitiveness and productivity. This is a story that the IT departments have a high stake in maintaining, since it's pretty much essential to establishing their position as a productivity centre and not a cost centre.

From the supply side, it's obvious that demand generation for new technology drives revenue, and accelerated obsolesence drives growth. But is this a good thing, either for suppliers or customers?

Capital life cycles have been getting shorter and shorter. For technology buyers, every deployment of a new generation costs the IT department in terms of operations, and the entire organizatin in terms of productivity. The faster the innovation comes, the more incremental each generation is, and the less overall productivity gain that is realized. And that's fine, since enterprises will self-regulate their pace of innovation anyway, whatever the hype tries to convince them to do.

But what are the negatives for the vendors and service providers? There's a case to be made that rapid innovation can cause great harm to companies and the industry overall. Once a new technology starts to be hyped, everyone needs to get into the game or get left behind. Venture dollars pour in, startups rush to market, and established players look for strategies to exploit or protect. And this whole cycle has been accelerating, so that a technology that used to take 10 years to play itself out now takes 3, and new product releases with the latest features and innovations have to get to market every 6-12 months. Product life cycles are shorter and shorter.

And so the industry becomes addicted to change. Like other addictions, it propogates itself in ways that are hard to break. The shorter the life cycle, the greater the pressure to release new innovations.

At the product level, code has to be written incredibly fast, and aside from the quality issues this creates, I've often seen it lead to bad deciscions that mean the entire code base of a product has to be rewritten every 2-3 years. Shorter life cycles means lower manufacturing quality, which I have to believe creates issues within the active service life of the product as well, leading to customer satisfaction and support cost problems.

At the business level, suppliers are under such pressure to innovate and keep up the pace, they become dependent on new revenue streams to survive, at the expense of efficiency and profitability. Tech businesses are very different from others in this respect, especially the smaller or younger companies. These companies need to sell the latest generation of product to survive, but (as I mentioned earlier) their customers aren't under the same financial or competitive pressure, unless enough of their competitors start realizing real competitive gains from the technology.

None of this is necessarily right or wrong, but I do think it pays to think about these larger issues now and then. The things we complain about in the high tech industry (like hype cycles, rapid change, too-slow adoption, etc.) have deep causes we need to understand and account for in our strategies.