In fact, the origins of every one of IBM's products can be traced back to breakthroughs in the company's research laboratories, says Paul Horn, the pony-tailed head of the famed Watson Research Center.
So how will Big Blue keep its technological edge as the basis of competition in the information technology industry shifts?
Selling services rather than hardware or software has become the biggest growth business in the computing industry - a fact that IBM was quicker to realise than most. Last year 48 per cent of revenues came from providing services, generating 41 per cent of its profits.
The difficulty is that services are just not susceptible to IBM's traditional way of handling technical innovation - what Mr Horn caricatures as "a geek inventing something in a basement". Those inventions have traditionally been packaged and sold to customers in the form of hardware - or, more recently, software.
There is not much science in the creation and provision of services: the universities do not recognise it as a separate area of study, says the IBM research chief. He compares it to the period before computer science was born some four decades ago.
How, then, can a company such as IBM hope to use its research and development prowess to maintain a competitive advantage in this new era of computing?
That question may have big implications for companies in all industries, not just in computing. As more value shifts towards services and away from products, the nature of competitive advantage changes. IBM, having blazed the trail towards services in the tech industry, faces the question earlier than most. Its research laboratories have already produced a number of early answers to this conundrum.
One is to take its scientists out of the "basement" and push them out into the real world. At the start of last year it announced plans to send some researchers out to work on consulting assignments - a programme known inside IBM as "rent-a-researcher". Besides putting more brainpower to work on complex IT tasks, it was also meant to stimulate new ideas that could be brought back and developed in the IBM labs.
Perhaps not surprisingly, this idea did not go down well with some researchers and it has been slower to get off the ground than the company had hoped. "Quite frankly, there was some apprehension at first," says Doug Dykeman, a manager in IBM's Zurich research lab, though he and other executives claim this early resistance has been overcome.
Last year 150 of IBM's 3,000 researchers spent at least part of their time working on consulting projects. Eventually that could rise to 300-400 a year, says Mr Horn.
A second response is to use research to provide an element of technological originality to a far broader services engagement.
For instance, IBM credits its success in winning a €60m (£40m) local government service contract in France to the cutting-edge research on security being carried out in its Zurich research lab. The contract involves the automation of the land records in the Alsace and Moselle regions. It was only after the customer had seen the latest ideas on security being developed in IBM's labs that it signed up for the broader contract, the company says.
Third, and perhaps more far-reaching, are the company's attempts to use research breakthroughs to devise new services that can be sold as high-priced consulting assignments.
Take a new service known as Web Fountain. Using algorithms developed in its research labs, this service digs far deeper into the internet than traditional search engines, for instance extracting information from chat rooms and blogs (journals available on the web). IBM has sold this capability - Mr Horn calls it "Google on steroids" - as a one-off service to companies that believe commercially important information lies buried in the internet. For instance, it has been used by BP to track the health of its brand, and by an unnamed record company to get an early warning about when and where demand for particular songs or artists will spring up.
In reality, this has as much to do with business model innovation as with technical innovation. What is significant is not so much the technological breakthrough itself as the way it is provided and paid for.
"It's not a pure software play, and it's not a pure services play," says Sharon Nunnes, an IBM executive whose job is to identify breakthroughs in its research labs that could form the basis of new businesses. "It doesn't fit into our current services model, but it's more than packaged software."
The new model combines hosting, in which IBM runs the service on its own computers, with high-end consulting. The company is also looking into whether Web Fountain can eventually be run on a customer's own computers. "We try to look at new business models, to look outside the box a bit," says Ms Nunnes.
In broad terms, ideas that are sold as specialised services today are likely one day to become standardised and sold to a much bigger market as products.
"A lot of things over time may become software products," says Mr Horn. While services such as Web Fountain demand close individual attention from research scientists at present, "over time you would expect these things to be cleaned up and included in our software," he says.
The key here is how fast that services-to-software shift can take place, and whether IBM has a vested interest in slowing the transition to maximise its own returns. Executives say they have every incentive to standardise the best ideas coming out of the labs to reach as big a market as possible, though Mr Horn adds of Web Fountain: "It will be a somewhat custom services engagement for quite a while."
With an eye on the longer-term development of the science that supports its services business, meanwhile, IBM says it is trying to stimulate academic thinking about the field. "It could be a fundamentally new discipline," says Mr Horn.
What that discipline will look like may already be taking shape in academia. In areas such as genomics and financial market analysis, university computer science departments are increasingly undertaking joint research that spans other disciplines.
"This is the beginning of services research," says Mr Horn. "It's about IT guys solving hard business and societal problems using technology." If he is right, it could spell a fundamental shift in the way IBM's research labs go about finding the next big breakthroughs in computing.
Richard Waters and John Gapper last week won the IFS Award for the Best Corporate Finance Submission at the Business Journalist of the Year awards
Big Blue breaks out of the box
By Richard Waters
Financial Times; Mar 17, 2004