There's one scary thing about making predictions. Someone might go back and compare your forecasts with what actually happened.
That's exactly what we did. We examined the predictions analysts made in two sectors - online advertising and consumer e-commerce - and ran the numbers against what official agencies say really happened.
How did the major industry analysts stack up? There's no simple answer. Researchers tended to underestimate online advertising revenue, which more than quadrupled from 1998 to 2000, according to the Internet Advertising Bureau. At the same time, many analysts were conservative when forecasting e-commerce sales - and that moderation helped them hit the mark after the market tanked in 2000.
The IAB says online advertising figures for 2000 will amount to between US$8 billion and $9 billion. (Final IAB numbers are due in April.) Most analysts forecasted low - in the $5 billion to $6 billion range. Only IDC's most recent prediction of $8.1 billion came close. What's odd is that most forecasts came in under the actual number - odd since critics often accuse analysts of padding their forecasts to cheerlead for emerging markets.
When it came to online shopping, most analysts over the past few years aimed their forecasts well below hype level. Then the market collapsed, dragging heaps of e-commerce merchants down with it. So when the final numbers were tallied - according to the Department of Commerce, Americans spent $25.8 billion online in 2000 - many analysts' predictions came in near the mark. Three firms - Jupiter Research, Gartner Group (IT) and Yankee Group - nearly hit the figure dead on.
Of course, which purchases you include under the consumer e-commerce umbrella makes a difference in the accuracy of the forecasts. For example, the US Commerce Department doesn't include travel, financial or other service revenues in its figures, nor does it include the amount spent on products researched online but purchased offline. Forrester Research includes these categories - therefore its predictions are higher. In past years, Yankee Group used a similarly broad definition of e-commerce but narrowed its definition last year to match that of the federal government. This helped Yankee nail its 2000 e-commerce forecast.
Analysts fancy themselves a sober lot. But all the new-economy hoopla last year seemed to make some forecasters downright giddy. From sampling snafus in Brazil to the hyping of events better left unhyped, some silly moments from 2000:
SEX AND SCIENCE
Researcher Alvin Cooper grabbed headlines when he published a report in psychology journal Sexual Addiction and Compulsivity concluding that the Internet encourages addictive sexual behavior, especially among women and gays. Turns out the survey had no scientific validity since the sample included only self-selected cybersex veterans filling out Web questionnaires. Cooper, who teaches at Stanford University and is the "Sexploration" columnist on MSNBC, defended the methodology.
ALERT THE MEDIA
While prepping for an IPO that was eventually canceled, Greenfield Online took to sending out press releases on increasingly trivial findings. Among them: Web users were "not particularly bugged by the flu"; thought The Sixth Sense should receive the Academy Award for best picture; and liked to post photos of their newborn babies online.
PHONING IT IN
Media Metrix's first findings from Brazil concluded that 25 percent of Brazilians owned PCs - and 75 percent of them dialed in to the Internet. But, skeptics quickly noted, only 12 percent of Brazilians have telephones. Media Metrix later said its figures applied to only the portion of homes with phones in 10 metropolitan areas.
WE ARE THE WORLD
3Com (COMS) , in partnership with Harris Interactive (HPOL) , spent millions on a global polling folly dubbed Planet Project Planet. The idea was to survey people around the world about everyday topics and post the findings instantly online. But because respondents weren't selected scientifically, the results were statistically meaningless. But that didn't stop the United Nations from endorsing the project - or 3Com from promoting findings on topics such as dreams, religion and sex. - M.T.