Nothing is as `vapoured' as the Net

Sometimes the "news" which comes from the Internet or is written about it seems more like a giant public relations exercise rather than fact-based reportage. Perhaps that suits a medium in which the companies most closely associated with it have skyrocketing stocks which seem, to earthbound observers, to have no visible means of support.

In Silicon Valley, the common term for a bucketload of hype with little of substance to support it is "vapourware". The term refers to heavily-promoted and publicised "future" products from companies which somehow never materialise, but the word also has become convenient shorthand for hype disguising hot air.

Nothing is as vapoured as the Internet. Take three things which should particularly be of interest to any company eyeing the Net as a business tool: estimates of the value of e-commerce, predictions of the numbers of people online, and estimates of the value of online advertising.

The numbers swing wildly from week to week, and are quoted endlessly by journalists (like me) to give context to a story. But who is coming up with these figures, and how? Occasionally a report is issued from an academic or research-based source, but most of the time the "analysts" who roll out such figures - often for the major firms most frequently relied on for information - are either people who have worked in PR or journalists moving into a better-paying field.

Rarely do they have specialist training in information research, nor are they statisticians. Or to put it another way, your insurance agent wouldn't use such figures to calculate the risks of even a minor policy. They certainly use more complex, sophisticated and proven statistical methods to determine your car insurance premium than Internet analysts use to produce figures on which entire industries make business decisions. Come to think of it, perhaps that's why stocks for companies which have never, or at best, barely, reported a profit are doing so well.

Thus, we get massively conflicting estimates for the future of e-commerce. Is it a $6 billion industry over the next few years? A $30 billion industry? Or even $300 billion? All have been offered by major industry analyst firms. And what are we estimating? Actual Net sales to end users, or do the figures include the value of business-to-business transactions which might have been conducted via email? If the latter, it's slightly duplicitous to offer the figure to promote the future of business-to-consumer commerce, yet this is regularly the case.

Another murky Web "statistic" is the number of people online. One report this week says 102 million; another 129 million, while another suggests 37 million. Where do these figures come from? Based on what expert knowledge? What does "online" mean? Five minutes' Internet use a week?

An hour, 10 hours? Context matters, and anyone who works in research can tell you that statistical "estimates" based on what seems likely are prone to significant errors.

At any rate, time spent online is of far more relevance in most cases than the number online. If 1,000 people enter the Stephen's Green Shopping Centre every Saturday, but only 300 of them stay around more than 10 minutes, the remaining 700 are of little real significance to the shopkeepers. That's what makes the 37 million estimate, by eMarketer, quite interesting.

To be "online" to it meant users who log on at least once a week, for at least an hour. The figure starts to give some meaningful context to Internet figures and also usefully damps down the endlessly unrestrained reports of the Web's commercial fabulousness.

Next, estimates of online advertising. In Geneva a few weeks ago a CNN online executive gushed at a major Web conference about the marvellous present, not to mention future, of online advertising. Reading a report in the Guardian, I wondered if I'd read that right.

I've yet to meet anyone who is trying to attract online advertising to a site who would agree with the CNN man's assertion and that includes some of the best "content" locations out on the Web, with reader demographics upmarket print magazines would kill for. It is interesting the glowing reports tend to come from those with a vested interest in attracting advertisers and assuring them that the Web is, well, fabulous.

Which brings us to a sobering article two weeks ago in the Silicon Valley newspaper, the San Jose Mercury-News. A reporter carefully researched, then shredded statistics from analysts on the value of Internet advertising. Most figures, the reporter argued, could be cut by two-thirds. It turns out that most ads are never strictly paid for but run as part of promotions or as benefit-in-kind or any of a number of other cashless or deeply discounted options.

Of course the Web is full of possibilities and potential and any business ignores it at its peril. But a wise business might find great use for a little scepticism. Perspective has more value than analytical vapourware.

Karlin Lillington is at klillington@irish-times.