My biggest mistake was to continue spending money on developing a product that was never going to work.
In 1984, when I was the European executive responsible for retail and banking product development, we were asked by the breweries to produce a special point-of-sale terminal for public houses. Bearing in mind the number of pubs and restaurants, it was potentially a hugely profitable operation.
The breweries outlined all their requirements, whereupon the product went into development, with about 50 people from IBM working on it full-time, plus various subcontractors.
But along the way, they lost sight completely of the problems they were trying to solve. Instead of looking at what was already in the market-place, they just kept adding bits and pieces.
Gradually the terminal became so big and so complicated that if a bunch of customers had come into the pub and asked to set up a tab it would have taken at least 10 minutes, by which time they'd have gone next door for their drink.
The fact is, people working behind a bar don't particularly want to fool around with a computer when they're frantically trying to serve. And since the majority work part-time, how would they be trained to use it? We should have realised this just by going into a pub, which, after all, we did quite often.
But we didn't really think about any of that. It was a classic case of only hearing what we wanted to hear. We thought people would adapt to the product we designed, which of course they won't.
We were only about a third of the way through when I first suspected we were making a terrible mistake. Everyone else was extremely enthusiastic, but I had this gut feeling that made me want to throw up every time I thought about it.
Someone's got to have the courage to say, 'I'll take responsibility for wasting pounds 10m. Come on guys, let's stop before we spend another pounds 10m,' and it should have been me, but that means you've got to admit your own mistake, which is hard.
It became increasingly difficult to back out, and instead of using common sense, I turned myself into a gambler. We always believed that the gold-mine was just around the corner, but when we got around the corner there was just another problem.
It was almost two years before the first one was installed in a pub, and we had to keep on spending to get it right. The customers got what they wanted in the end, but IBM paid the price.
Of course, once you've got that far, then what do you do? You want some natural disaster to occur, because really you want someone else to make the decision for you. We never did sell more than a few hundred terminals.
That's why we're passionately restructuring at the moment; to make sure nothing like this can happen again.
We have changed the system so that there is always one person directly responsible for each product. And we form alliances with companies in touch with those niche markets.
It's easy to get carried away when you're spending someone else's money, but that's another thing I've changed. If someone wants pounds 10m for development and I think it's a good investment, I'll give it to them . . . but I'll charge them interest.
These days, I act more like a banker, and by imposing these disciplines, we are flushing out opportunities for waste.
(Photograph omitted)Reuse content