Beware of Optimists

Don't base your business decisions on overly upbeat market research

I'm no soothsayer, but I had a hunch the online grocery business would spoil like a carton of outdated milk. So when a buddy told me he planned to give up his job with software giant Oracle Corp. (ORCL ) for a post at Internet grocery startup Webvan Group Inc. (WBVN ) a couple of years back, I scoffed that IPO mania had turned his brain to cottage cheese.

By last July, Webvan's wheels had come off, and my buddy was skimming the classifieds. He didn't see it coming, but then neither did a lot of folks. Among the most surprised must be researchers at Jupiter Media Metrix. They said online groceries would be a $2 billion business last year--more than twice the $800 million reality. Before you lay too much blame on Jupiter, though, remember that most analysts once sang the praises of ill-fated online endeavors such as selling postage stamps and pet supplies.

It's a bit odd: The tech industry can make supercomputers that handle enough complex calculations to predict the weather but can't produce market data that holds up much longer than a thunderstorm. When forecasting tech market growth, analysts span the map like a bunch of lost tourists. EMarketer Inc., a researcher that evaluates Net forecasts, checked predictions for high-speed Net access that 16 analysts made in 2000 for 2001. Only six got within 10% of the 11 million U.S. homes that had broadband by yearend. London-based Ovum Ltd. apparently threw a dart and missed, predicting 15 million users. Again, I'm not clairvoyant, but anyone who has tried broadband knows that at $50 a month it costs too much. Plus, it takes the tech skills of an engineer to get it working.

QUESTION THE HYPE.

  What gives? The worst forecasts are made as a market's hype-to-reality ratio is at its peak. In 1999 and early 2000, when the world was buzzing about the promise of Web media, most researchers guessed that banner ad sales would explode. AdZone Research Inc. estimated that online ad sales would hit $16 billion last year. But banners bombed: Just 0.3% of Web surfers click on them. So ad revenue finished last year at $7.3 billion--10% lighter than in 2000.

I asked the researchers for an explanation. They all say they try to account for economic slowdowns, funding shortfalls, and demand shifts. Obviously, no one could have predicted September 11, and seeing the depth of last year's tech downturn wasn't easy. But researchers need to question the hype companies feed them. Ovum, for instance, says its predictions were colored by telecom companies' ambitious development plans. "We sort of cheerfully assumed [managers] knew what they were talking about," says analyst Tim Johnson. And Jupiter analyst Ken Cassar says he based his forecast of grocery sales in part on Webvan's predictions of rapid expansion. "The market didn't develop nearly as quickly as we thought," Cassar says.

That's an understatement. Jupiter now says online grocery sales this year will be $1.2 billion, a shadow of the $4.2 billion predicted two years ago. The lesson here is that you need to be ready for both boom and bust. Cost-conscious companies such as shopping site BizRate.com are still around because they didn't get suckered by upbeat forecasts. "As a kid, I saved my Halloween candy for the entire year," says BizRate CEO Chuck Davis. "In a Net business the same principle applies: Save the money for a rainy day." If market researchers and Webvan's managers hadn't bought each other's hype, the online grocer might still have its wheels and my buddy might still be rolling along with it.

By Roger O. Crockett

    Before it's here, it's on the Bloomberg Terminal.
    LEARN MORE