Don't buy the hype in wireless research
[Ziff-Davis Wire]
e-topics.com
[ Another person skeptical about wireless forecasts. However, I dont think his 523 Rule is the solution - DPR ]
Enterprise via NewsEdge Corporation : What do goat entrails and chicken bones have to do with wireless? More than they should, I'm afraid. In ancient times, augurs predicted the future by searching for clues in fresh entrails or tossing chicken bones. Today, we rightly scoff at such prognostication methods, but otherwise reasonable and knowledgeable people still swallow predictions that, but for a thin patina of authoritative research, might as well have been produced the old-fashioned way.
Take the wireless arena, which analysts have pegged as the latest field of dreams. While firms vary on the conservative/optimistic scale--IDC sits close to the conservative end, whereas Jupiter bookends the field at the optimistic end--most wireless research reports strike me as so wildly hopeful that I almost expect Shoeless Joe Jackson to part the corn stalks in the outfield and saunter into view.
Why should you care what these reports say? Well, in my experience, CEOs and other executives often use the projections contained in these reports to prod their IT managers. Therefore, by staying on top of expected trends and projections, you'll be able to educate top management on what can be realistically accomplished--especially in the nascent wireless arena.
Where's the method?
To see beyond the hype, examine the methodology behind a research report--if you can find it. The brief recitations of a firm's analysis in articles or elsewhere rarely link to the report's methodology or possible margin of error, nor do they reference the accuracy of the firm's past reports or give an estimated range (i.e., there will be 30 million to 50 million wireless users by 2005 in the U.S.).
It takes some digging through most research companies' Web sites to uncover any understanding of the methodology that results in their sage figures. Jupiter is fairly forthcoming by prominently listing a methodology page in its research section. Gartner does the same, and includes a solid breakdown on its probability scale. IDC, on the other hand, does not provide any explanation of its report methodology.
Sadly, research firms often says as little as possible about how they come up with their numbers unless you sign up to be a client, or they supply just enough of an outline for you to know a defined methodology is at work, but without any firm details.
We've all been conditioned by a generation of commercials toting one-line statistics to believe that "four out five dentists" recommend such and such, without ever questioning which dentists, how many, and how they were asked. The result is that when a one-liner slides down the chute, such as Jupiter Media Metrix's latest prediction that "U.S. wireless Web users will increase from 4.1 million in 2000 to 96.0 million in 2005," it garners much more weight than it deserves, despite the many caveats Jupiter includes in the full report.
Cut them down to size
You won't be misled by overoptimistic analysts if you follow my 523 Rule when factoring market research into your wireless plans. Here's how it works. Unless a research firm has accurately tracked a market for five years, take its current analysis and double the length of time predicted, then cut the size of the market to one-third. For example, Jupiter predicts there will be 96 million wireless Internet users in the U.S. by 2005. My 523 Rule transforms this into 32 million wireless Internet users by 2009.
What else can you do to protect yourself and save your company from embarking on wireless initiatives that could turn into boondoggles? First, focus on what's already working. 3G, a third-generation, high-speed wireless system that extends Global Packet Radio System data to provide peak speeds of 384Kbps, isn't going to be widely available for five years, at the earliest, and, then, only after a massive round of bankruptcies and consolidations ripple through the market.
On the other hand, Wi-Fi, also known as 802.11b wireless networking, running at 11Mbps, is real, relatively inexpensive, and should definitely be on the roll-out schedule of every up-to-speed IT department that's considering pulling new cable in the near term. I use Wi-Fi every day, all day, and it's a godsend. (I'll go into this in greater detail in a future column.)
My only wish is that Metricom and an 802.11b vendor (hello, Sierra Wireless, are you listening?) would produce a dual Ricochet 128Kbps/802.11b PC Card that would provide continuous wireless coverage (in major cities) both inside and outside the enterprise.
Do you believe what you read about the future of wireless? Talk Back below. Rich Santalesa is the chief analyst at PDA & Wireless World, an editorial, network, and wireless analysis firm based in New York.
<<Enterprise -- 03-14-01>> |
