Two years ago, when I started at Dobedo and people asked me how the site was going to make money, I replied, genuinely, "We don't need to make money yet."
I had been told by my boss that the capital backing the company would last as long as we needed it, so we could concentrate on increasing traffic and building brand awareness. "If this company fails, it won't be due to lack of money," he told me. They were prophetic words.
We began a full European roll-out without ever really knowing how we were going to make money. We were "testing" different revenue models. If we could capture the attention of enough people in the 16-24 market segment - and in Sweden, 25% of 16-24 year-olds were registered Dobedo members - then the money would surely follow.
Wrong. By the time we had worked out how to make money out of users - and we did work it out in the end - it was too late.
Martin Osterdahl, who was in charge of the international roll-out of Dobedo, is now North American area manager of Biondi Industries, a company specialising in making early investments in fledgling digital companies. He has learned much from his experiences.
"Before you even think of a name for your dotcom, nail three sheets of paper on your wall stating your three independent and sustainable revenue sources and their respective competitive advantages," he says.
"Look at them. If they make sense, take them down and put them in a drawer. Do something completely different for two months. Repeat the procedure. If they still make sense, then try convincing top talent to join you, without having secured millions in financing."
Dotcoms love to boast about their registration, page views and unique visitor statistics. At Dobedo these stats led us to begin to believe our own hype. A lot of other people believed it too. One industry publication proclaimed us "the future of youth marketing" but the truth was less fantastic.
It's easy to get carried away. It's easy to miss the stats that show what's really happening with site traffic, as Stephen Spencer, an associate director at NFO-BJM Research specialising in research for dotcoms, points out. "It sounds obvious, but many Web sites do not really know who their visitors are," he says.
"If you have a registration-based site, it's really important to know why people who come don't register and become regular visitors. One famous e-commerce site regularly boasted how many customers it had, but it had around five times that number who came and did not register. Why? They never bothered finding out. The site went down the pan last year."
Venture capital is not always the god-send it appears to be. Osterdahl warns: "Avoid venture capital money and dependence as long as possible. And don't take for granted that financiers, business partners, staff or board members have the same agenda as you, just because you're involved in the same project."
Although often guilty of naivety, many dotcoms were not helped by the way venture capitalists threw money at almost any commercial online venture but failed to back it up with sound business advice and help. Inexperienced entrepreneurs have seen their businesses go to the wall, despite the fact that their ideas were commercially sound, because they were not given enough help by the more experienced businesses that were backing them.
The fate of Boo.com, which pioneered new Flash releases, illustrates perfectly what happens when you build your business on technology that is so "state of the art" that no one else is using it yet.
But dotcoms continue to make the mistake of using their prospective customers as technical guinea pigs, without attending to the fundamentals - a site should be fast, easy to use and accessible 100% of the time.
Mark Blackstock, head of telegraph.money.co.uk, explains, 'Technology can make many of our dreams come true, and we love it for that, but we shouldn't believe that all of the time.
"At Dobedo we launched a 'new and improved' version of the site, using cutting edge Broadvision database technology, without ever fully testing it in a live environment. The intention was to enhance the potential for
e-commerce functions on the site. But the site didn't work for over two months. It took over six months for traffic figures to recover, by which time our financing had run out." It seems so obvious, yet the complacency among many dotcom entrepreneurs when it comes to the fulfilment of service remains, on occasion, staggering.
At Dobedo we boasted that our users were the most loyal on the Web, and it was true that the regulars displayed high levels of enthusiasm for the site. But even the die-hards eventually gave up in the face of prolonged technical hitches and difficulties accessing the site.
The "facelessness" of the medium has led to many companies falling into the trap of thinking that an online business does not have to worry about personal service
Spencer says it is important for dotcoms to remember that they are not exempt from high customer expectations when it comes to delivering high-quality service and goods.
"Just because people like your site, like your products and are buying from you doesn't guarantee success," he says. "Our research has shown that if products take four weeks to arrive, customers won't come back. And it's a much more common problem among dotcoms than you might imagine."