Reality Checks and Balances

Even the most technically appealing products end up as dead ends. To avoid them, you need to stay tuned to market realities.

• By Laura Wonnacott
• 11/01/2001

There's no doubt that our industry comes up with some real show-stoppers, both good and bad—and manages to do so in rapid succession. The question is: How could you have avoided the money-wasting dead ends? And how could you have guessed earlier when it was time to jump on some of the really good stuff that has surfaced in the last 10 years?

In many cases, as with products and ideas like OS/2 or WAIS (Wide Area Internet Search), a failure isn't an indictment of the technology or product itself. Many were very good, often superior to the competition that ultimately finished them off. For a recent example, look at Metricom's Ricochet, a high-speed wireless solution often viewed as superior to cell-based technologies like CDMA. What doomed Ricochet was the need to build an entirely new infrastructure; CDMA-type solutions can piggyback on the existing cell infrastructure. (By the way, there's still hope—last I heard, Ricochet may have a buyer.)

The real world is about a lot more than technical merit. Regardless of the underlying cause, a dead technology into which you've invested company money means wasted resources. You as a technical leader need to help top management stay grounded in the political and market realities of the times. That, as much as the actual merit of a particular idea, is what propels or dooms a technology.

So how do you see into the future? An easy way to avoid technological dead-ends, of course, is to wait until it's clear who's won. But that dooms you to following the leaders around, waiting for them to let you know which new technologies have long-term potential. By the time that's clear, your competitors will be firmly entrenched and far ahead of you. Instead, you need to strike a healthy balance between reckless investment and fearful lagging.

You need criteria for technical investment based on the industry your company competes in. You also need to take a creative look at the risk/return ratio for each new technology. If a product or technology has a slim chance of paying off big, but won't cost much, why not try it? Small risk, big possible return. Examples are things like paying search engine placement firms or renting online mailing lists. Sure, either technology may not be worth anything eventually, but for the time being, they're pretty safe—and cheap—bets.

You can probably calculate the eventual cost of a new idea fairly easily, but figuring the potential payoff is a bit trickier. Try thinking of analogs in your own experience or history. For instance, I would argue that the potential payoff for WAP or other text-based phone browsing compared to traditional HTML is something like comparing Cliffs Notes to a good book. It's a workable solution if that's all you've got time for, but anyone with access to both would clearly choose the fuller version.

Finally, you need to figure out the likelihood of the risk of failure. This is much more difficult—actually more of a black art than a science. And you're going to be wrong sometimes; there's no escaping that. But calculating risk is an interesting thought experiment that can be educational—and just going through the exercise can help put technology investments in perspective.

You'll want to consider the technical hardiness of the issue: Is it first-generation, or a refinement of an existing product, protocol or standard? How widespread is deployment, and are there any numbers on growth? Who are the main competitors? What market forces might support or oppose the protocol? For instance, any product or technology that is in direct competition with Microsoft, especially if it's the only competitor, has a drastically lower chance of survival.

Remember—there's no universal process for evaluating any technology under consideration. It's impossible to foretell the future, though it's very possible to rationalize the decision-making process. Through experience and plenty of painful lessons, your own model will continue to evolve. Hey, you may never achieve prescience, but you can at least look like you know what you're doing.