A funny thing happened during the writing of this column. I became a millionaire. In true new-economy style, this bore almost no relation to hard work, a flash of genius or good luck. It had every possible relation to febrile markets.

Before the taxman picks up the phone, I should make it clear that my millions were strictly in Turkish lire. My elevation to the millionaires' club lasted precisely as long as my weekend trip to Istanbul. As the exchange rate at the time was 1.3m lire to the pound, we lira millionaires are somewhat less than a select group.After a heavy devaluation of their currency, Turks today are like the dotcommers of 18 months ago: staring at huge numbers on paper which seem to have lost touch entirely with reality.

If the centuries-old world of currency trading can succumb to the lure of crazy numbers, then it is hardly surprising that the immature e-commerce sector should also prove susceptible. Then again, number inflation has characterised the sector from its earliest days.It was the exponential growth in total online users which first attracted venture capitalists and others to commercialise a medium hitherto dominated by academics, technologists and niche community groups.

In every successive wave of internet technology, quantity has been privileged over quality. Thus internet service providers fought to sign up more customers, even if this meant offering non-viable pricing packages or sacrificing customer service. Advertising-supported websites have developed a whole gamut of reporting statistics for their clients - from hits, page impressions, click-throughs to unique users - in order to count so-called "eyeballs". When practically unknown websites can claim the same number of users as established magazines or even television channels, and in the absence of agreed ways of measuring online audiences, observers are entitled to be sceptical.

And so the argument goes. E-commerce will be bigger than auctions. Mobile will be bigger than e-commerce. B2B will be bigger than both - and therefore better. (Just look at the charts.) Numbers, numbers everywhere! How many doomed dotcoms persuaded investors to part with their money by brandishing a couple of website logs, an opinion poll and some over-optimistic research, with little or no empirical data? And how many became doomed because they sought earnings at any price, rather than quality revenues achieved at viable margins?

To put it bluntly, no one has yet produced even a convincing formula for calculating how much a single ISP subscriber might actually be worth. So anything more ambitious is likely to be "finger in the wind" type forecasting. The irony is that now that western Europe is facing credible huge numbers - such as the cost of re-training workers in technology skills, delivering e-government or providing high bandwidth - the desire for investment has flagged.

Is it too old-fashioned to say that doing some or all of these things would be justified not because they would deliver some large statistical saving, but because they might improve the quality of citizens' lives and their experience of public services? You can't put numbers on those. That doesn't mean they don't count.

E-mail Carols Gande at carlos.grande@ft.com