Autonomy's speciality is analysing big data — using probability theory to find patterns across multiple different stores of information. It is ironic, then, that HP apparently missed the claimed massive accounting irregularities in Autonomy's own books — signals strong enough in retrospect to make the British company's actual worth barely one-tenth of the $11bn acquisition price.
The real mystery, though, isn't what happened or why — although one must wonder whether some rogue chemist has been spiking the coffee in HP's boardroom, such is the unreality of the chain of events. HP's board anointed Leo Apotheker without an interview, voted through the Autonomy deal despite the severe reservations of HP's own CFO, paid $11bn for a company with assets of around $3bn, and then blamed that overpayment on previously invisible yet poorly defined accounting misdemeanours.
HP's one detailed complaint about the accounting, that Autonomy sometimes booked hardware sales as software to inflate the margins, only accounts for some $200m — hardly chicken-feed, but hardly a multibillion-dollar heist either.
Things didn't go as expected from the moment the deal was signed. In May, the Financial Times noted that Autonomy's revenues dropped off a cliff the moment HP took over — which people within Autonomy claimed was due to a difference in corporate culture, but the FT noted dryly may have been more due to a disagreement over what constitutes a sale. There is some flexibility, sometimes over-exuberantly used, over when expected revenues from a deal are actually put on the books.
In 2011, the FT noted that Autonomy was aggressively buying up smaller companies for more money than might seem sensible. It also pointed out that Peel Hunt analyst Paul Morland had said: "This appears to be consistent with our theory that 2011 forecasts were going to be hard to achieve and the acquisition fills a hole."
Two years before that, there were reports of overstated revenues and payments deferred — more useful tools to make figures look healthier than perhaps they were.
There are plenty of discussions about where this sort of behaviour lies on the spectrum of accounting, from conservative to over-inventive. Too much can land a company in regulatory hot water, or worse. But the key fact here is that if you can read acerbic reservations about a company in the FT over many years, so can any one of HP's 15 different sets of financial advisors in on the deal.
You'd hope that some of them had a subscription to the newspaper; they could afford one — a year's worth of online access to the FT clocks in at around £350 ($560), and HP's collection of banks, accountants and dealmakers split some £44m ($70m) in fees between them. (It's not clear how much those fees would have been reduced had the deal not gone through.)
This wouldn't be the first huge deal to go through for the wrong reasons, with the wrong people, at the wrong time. But the size and specifics of the calamitous Autonomy purchase make it special.
And that's at the heart of the real mystery — how on earth can HP now present itself as a company that understands either software or business? The company built its reputation on hardware, and has never really been at home in enterprise software.
How on earth can HP now present itself as a company that understands either software or business?
It lacks a complete stack to sell to CIOs, who increasingly inhabit a hardware-free world, and it lacks a coherent strategy to explain how this will change. And any claims on its behalf to have special insight into the strategic needs of the enterprise now have to be weighed up against its completely fumbling its own.
HP's future is bleak. Revenues are down across the board, in both hardware and services. Its natural habitats — the server room and the enterprise desktop — are shrinking rapidly, as mobile and cloud lay waste those once-fertile rainforests of IT spending. HP is talked of as a take-over target, even as fellow-travellers such as Intel and Dell are looking shakier than ever before.
Autonomy may have been a shocking failure, but it's not the only one. Most worryingly, it seems just another indication of an HP that's lost its way in dangerous times, with CEO Meg Whitman running out of cash, choices and sympathy. It's past time to stop pinning the blame and start playing the game of making HP coherent, relevant and believable.