The Autonomy founder’s relationship with HP hasn’t worked out – a familiar tale of eccentric inventors and big companies
The news broke today that among the 27,000 workers to be shed by Hewlett-Packard is Mike Lynch – the founding chief executive of Autonomy, the software company HP bought in October 2011 for $10bn – and many of the employees he brought with him.
Lost in those numbers is the fact that Lynch is a fascinating and original thinker. It’s little surprise, in a way, that he and American corporate culture didn’t find a match. Autonomy began its life in 1991, when Lynch borrowed £2,000 to start Cambridge Neurodynamics. Five years later, believing there was more scope in Autonomy, that company was spun off and Lynch went with it. His key insight was that the theories of the then little-known 18th-century Presbyterian minister and mathematician Thomas Bayes could be used to search and understand unstructured data. The company stumbled a bit at first, attempting to implement these ideas on PCs. It was when Autonomy’s software moved to the server side of the equation that the company began to take off.
Today, Bayes is everywhere in the computing world – the American author Sharon Bertsch McGrayne made just this point in her 2011 biography of Bayes, The Theory That Would Not Die. But in 2000, when Autonomy had become a billion-pound company and I interviewed Lynch for Salon.com, these ideas seemed exotic and downright weird. This was even more true when Lynch, despite his engineering background, talked about perception instead of software and said that “the area I work in is pure philosophy”. It took McGrayne’s book to completely understand what Lynch meant: Bayes provides a method for reasoning backwards, starting from an effect to understand what its causes might have been.
Had HP read that interview, the company might have grasped how much of a mismatch the two companies were likely to be. It’s not just that Lynch was used to being in charge himself; it’s that a Silicon Valley company with IBM envy is an unlikely culture to understand a slightly eccentric British thinker.
Of course, some relationships between eccentric inventors and large companies do seem to work out. Jeremy Ruston (the founder of Osmosoft and creator of Tiddlywiki) and Bruce Schneier (the security expert and founder of Counterpane) are part of BT. BT’s research lab, when it was under the charge of Peter Cochrane, was once home to an ant farm (to study communications). And several years ago, IBM picked up Jeff Jonas, whose thinking on data analytics is unlike anyone else’s. IBM was notorious for decades as the world’s stuffiest company, but its research division has always been a dramatic exception.
HP in the 2000s and 2010s is not the same kind of company, despite the legendary status it long enjoyed in Silicon Valley. For many, it is “company zero”: William Hewlett and David Packard had the first start-up-in-a-garage and the first big success in electronics, and its founders were widely admired and respected. Things are different now, between the public squabbles over acquiring Compaq and a spying scandal. The $1.2bn purchase of Palm in 2010 was supposed to take the company into smartphones and tablets. Less than a year and a half alter, the company exited that business entirely. Even more disruptive, the company has had four CEOs in just six years; the latest of them, former eBay CEO Meg Whitman, took over last September, weeks after the Autonomy acquisition went through.
But HP’s loss is Britain’s gain. Someone like Lynch is going to have more than one big idea. I can’t wait to see what he comes up with next.
• Follow Comment is free on Twitter @commentisfree