'When I found out that the software didn't work I began to get palpitations'

Trevor Burke, 48, has built PhoneLink into an pounds 82m electronic information services company. But not all his ventures have worked out so successfully; he unwittingly sold his first one for a song

Even after I had been in business for several years, I was still incredibly commercially naive. I sold my first serious company in 1987 for pounds 250,000, a price/earnings ratio of just one. It wasn't until more than two years later, when I was expanding PhoneLink, that I discovered the purchaser had pre-sold my company to a buyer in Atlanta for pounds 10m, 40 times what he gave me for it. I understood engineering, but I didn't understand value.

Being in business is a process of constantly learning lessons like that. I wasn't a very good academic student; I enjoyed the social side of being at a polytechnic too much. But after I left college I realised that part one of the guaranteed success story was to move around a lot. In order to really make money you need a great deal of experience across a lot of fields.

Even as a child I liked taking things apart and reassembling them, and there were never any pieces left over, so it was natural that I should go into engineering. My first job was designing heating and ventilation systems. From there I moved into sales, followed by my first spell as a director at a small family firm.

While I was there I ran a burglar-alarm business on the side, mostly to get some experience running my own firm. I also built my first computer from a kit. Once I got it running I started modifying it and taught myself to program. My first commercial product was a heat-loss program. I left to set up my own company, Technique, and shortly afterwards I was asked by an acquaintance, David Spicer, to help him with his kitchen-design software. By the time I sold the business it had a four-year lead on the competition in computer-aided design.

There were a lot of financial imperatives on me in those days. I had a new wife, four kids in fee-paying schools and no carpets on the floors. We were working flat out, sometimes until 2am then starting again at 5am. The effect on my health was staggering.

There were several near disasters during the early years. We worked for ages on an electric-analysis program that would have helped to design safe wiring systems for houses and offices. We were writing in a language called Basic, which was too slow, but Microsoft was promising to bring out a new compiler (a piece of software that translates human programming languages into the ones and zeros of machine code), which would have made our program run faster. We were also waiting for the government to introduce a new set of standards. When I found out, on the same day, that the government had backed down and that Microsoft's compiler didn't work, I started to get palpitations.

We also had the usual cash flow problems that most small businesses face. On one occasion a major London-based kitchen design company asked us to provide it with our computer system on a trial basis. Shortly afterwards it went into receivership owing us pounds 30,000. My wife and I spent the whole weekend driving to all its offices that were still open and forcing the staff to give us our kit back.

I know far more about starting companies with no cash than I do about starting them with cash. At one point in 1982 I had to ask the staff to go on half wages. They were very supportive, but in retrospect it would have been better to let half of them go. It put us all under tremendous stress.

Two years later, on Christmas Eve, the bank refused to pay my staff's wages at all. The company's overdraft was still reasonable, but when my banker added in the personal overdrafts it was over the limit and he panicked.

I almost sold one of the businesses then but my solicitor refused to do it. He said it was a minor bank problem and I should open an account with a different bank. He was right. You don't often get advice like that from lawyers.

While I was running Technique, my wife, Heather, and I set up another company which she managed. Profile UK specialised in getting information on planning applications to our clients. We produced 3,500 customised magazines a fortnight for them. We wanted to add relevant phone numbers to the magazines, but the cost was incredible. That was when I realised that the real cost of getting information was labour. No matter how good the database, it still takes time to retrieve it. And it wasn't in the interests of the people offering the service to make it go faster. That was the niche we aimed to fill when we started PhoneLink.

Like many small businessmen I tended to distrust fee-seeking advisers. But about four years ago I was sitting with a friend from Ernst & Young, the accountants, who had been trying to get me to bring in professional advisers. He told me about a chip shop he was working with that was going to make Harry Ramsden look tiny. I discussed what I was planning with him and the whole world changed. Now we're aiming to replace the Internet, which I think is fundamentally flawed.