Thursday, February 7, 2013

Will HP's accusation that Autonomy cooked its books stand?

This past November Hewlett Packard (HP,) a $130 billion company, and the largest manufacturer of PCs in the world, announced that it was writing down $8 billion of the $10.3 billion it paid to purchase Autonomy, a British software company. HP’s spokesperson argued that Autonomy had engaged in "serious accounting improprieties" and "outright misrepresentations” that created an unrealistic picture of the company’s worth. It linked these accounting improprieties to $5 billion of the $8 billion write-down. Michael Lynch, Autonomy’s co-founder and CEO, disputed the charge, and noted that HP had not provided any details justifying the accusation, or the size of the write down. As one report notes, “Mr. Lynch has come out swinging, denouncing HP's assertions as ‘completely and utterly wrong.’” Moreover, Autonomy’s British auditors, who had earlier investigated an employee’s charge of such improprieties, found no evidence for them.

While it is difficult at this point to surmise which party is telling the truth, or more realistically, where parts of the truth may lie, it is clear that Autonomy is a going concern with a thriving business. It produces software that enables companies to search structured and unstructured data, for example data from videos, radio programs, voicemails and text. The company is part of the emerging world of “big data” through which companies and government agencies can “scrape” and analyze very large data sets, for example in tracking the online behavior of millions of consumers. At the time of its acquisition, Autonomy had over 400 OEM partners, more than 400 vendors and integrators, and over 20,000 customers, including BP, Halliburton, Royal Dutch Shell, BAE Systems, AstraZeneca, United States Department of Defense, United States Department of Homeland Security, Citigroup and Symantec. Moreover, in January of this year, HP announced “a significant customer win” when NASCAR, the premier stock car auto racing organization in the United States, agreed to use Autonomy’s software “to power a Fan and Media Engagement Center. The software will be used to monitor print, radio, video, Web and social network sources and analyze fan sentiments on races, drivers, sponsors and all NASCAR activities.”

I believe that Autonomy did not perpetrate fraud as did Enron or WorldCom, but I think it is clear that Autonomy did engage in aggressive accounting practices. As one report notes, “The firm recognized revenue upfront that under U.S. accounting rules would have been deferred.” It also struck "round-trip transactions”—deals where Autonomy agreed to buy a client's products or services, while at the same time the client purchased Autonomy software. If for example, Autonomy sold company X, $1 million worth of software, by promising to buy $.5 million of services in return from X, the net impact on HP's revenue should be positive $.5 million. Yet in one round trip transaction, with a software company called VMS, Autonomy booked the cost of the VMS software it purchased, as a marketing expense -- it used the software to make marketing presentations --rather than as the cost of the software that Autonomy originally sold to VMS. But when analysts look at the intrinsic profitability of a business, they focus on the direct expenses of producing a service, “the cost of goods sold,” while ignoring overhead items such as marketing expenses. HP's accounting tactic exaggerated the profitability of its original sale of software to VMS. 

An alternative hypothesis for explaining the write-down is that HP overpaid for its purchase. It is common knowledge for example, that when a company announces its intent to acquire another firm, the stock price of the latter goes up while the stock price of the former goes down. Overpaying is expected. It is also likely, as so often happens, that HP found it difficult to integrate Autonomy into its operations. Discussing the latter’s performance some 18 months after its purchase, HP’s CFO told analysts, "License revenue was disappointing, sales execution was a challenge, and big deals were taking longer to close."

One reason for this disappointing performance may be that the co-founder and CEO of Autonomy, Mike Lynch, described as a “domineering” figure, was very hard driving and had created a great deal of pressure for sales. As one sales person reported to a newspaper, “Autonomy (prior to its acquisition) used a homegrown computer program, called SMS, to record salespeople's contacts with potential clients and calculate their rates of closing sales. Executives would sometimes use the data to craft broadly distributed emails calling out under-performing salespeople,  Those whose rates fell below a standard deviation  would be criticized or fired, the former salesperson says.”

Another news report noted that at one point, “Autonomy stocked piranhas for a while in the office fish tank.” Of course, this may be folklore, but even as folklore it highlights peoples’ perception of the setting as ruthless.  It makes sense, in this context, that once HP bought the company, and Lynch became a divisional executive with delimited power, the pressure for performance he stimulated, diminished as well. This is a common problem when large companies buy owner or founder-led businesses. Entrepreneurs are motivated by both the prospects of great success and the fear of catastrophic failure. Once the prospect of “hunger” is banished through access to the acquiring company’s resources, people feel less hungry to succeed.

The Autonomy purchase and its subsequent write-down, while big news, was overshadowed by a more important event that nonetheless gives added meaning to the purchase itself.  When the then HP CEO, Leo Apotheker, announced his company’s purchase of Autonomy, he simultaneously told reporters that HP would be selling its PC division, which on a stand-alone basis would have been the biggest personal computer manufacturer in the world.

This was riveting news, particularly since, when in 2002, HP first bought Compaq Computer to create its PC division, there was considerable bloodletting, leading the HP board to sack Carly Fiorina, the charismatic CEO, who pushed for the purchase. Ten years later, the PC division accounted for one-third of HP’s revenue and one-fifth  of its profits. As a result, the day after Apotheker announced Autonomy’s purchase, and the sale of its PC division, HP’s stock price fell 20%, eliminating some $13 billion in value. The board promptly fired Apotheker.

In retrospect, it appears that Apotheker’s decision to sell the PC division was part of a larger strategy to move from the world of hardware to software, much as IBM had done when it sold its PC division to Lenovo. Autonomy’s purchase, I propose, was an incidental tactic to underline this shift. Strikingly, as the Wall Street Journal reports, Apotheker tried to buy two other software companies both of which the board rejected, before agreeing to his bid to buy Autonomy. As the Journal recounts the story, Apotheker, in advocating for Autonomy’s purchase told Ray Lane, the chairman of the board, "I am running out of software companies to buy.” As this turn of phrase suggests, he was facing time pressure to underline his strategy of moving to software, and needed to buy, as the saying goes, “something, anything.” 

This suggests that one additional reason for the write down was that Apotheker and the board, facing time pressure and scarce opportunities, overpaid for what felt like a “last chance” acquisition.  Autonomy in this sense was valuable, less for the foothold it provided in the world of software, though its technology was certainly sound, and more as an object of desire to signal both to the market and to powerful HP’s executives, that the company was moving irrevocably into the world of software. 

Indeed, the executives of large corporations sometimes face insuperable obstacles when they try to transform companies from within. They face guerrilla warfare, as divisional executives, for example, in the case of HP, the executive team in charge of the powerful printer division, resist the transformation. Sometimes the only recourse is to buy and sell whole divisions, in this way shedding obligations, commitments and the political coalitions associated with them. This may be one reason that Apotheker and the HP board were willing to pay a premium for a company that was already highly valued by the market relative to its actual earnings. As one report notes, “The markup, about a 50% premium over the stock at the time, wasn't unusual for a software deal, but Autonomy's valuation was already on the high side. Its market value of about $6 billion was seven times annual revenue and 15 times operating profit.”

Was Apotheker’s strategy of moving to software sound?  One stimulus for the strategy was the effort, begun before Apotheker became CEO, to build a new hardware ecosystem based on a proprietary tablet computer and an operating system, WebOS, purchased from Palm, the failed tablet company that could not compete with multifunctional smart phones. HP saw this strategy as a way to compete in the emerging “post-PC” world in which tablets rather than desktops took center stage. Yet the tablet, called Touchpad, sold poorly and Apotheker shut it down. At the same time, the margins HP earned on its PCs were small, in the range of 5%, while, as the following table shows, the number of PCs it shipped had been slipping. Indeed, this shrinking market is one reason that Dell Computer is going private. It needs to shield itself from the stock market that is already discounting the future prospects of the PC. 


In short, after HP failed to reposition itself as a tablet manufacturer with a proprietary operating system, Apotheker was confronted with the task of selling a low margin product, the PC, in what looked like a shrinking market. It seems reasonable in this context that HP’s next step was to move fully into the world of software. Indeed, when the board fired Apotheker and appointed Meg Whitman in his place, Ray Lane, the board chair, noted that, as the one time head of E-Bay, Whitman had been a buyer of technology on a large scale, and was very familiar with “unstructured data,” the sweet spot for Autonomy’s products and services. The same month she was appointed, Whitman herself said, “That the company under her leadership would stay the course and continue to transition to an enterprise software business.”

So why fire Apotheker? One fact is clear, HP’s customers were very upset with Apotheker’s announcement that HP would sell its PC division. This was one reason for the drastic drop in the stock price the next day. Many customers were wholly vested in HP’s products and services, buying PC’s, printers, servers, and networking equipment from the company. HP’s PC product line, it turns out, was a strategic asset stimulating the sale of other equipment as customers bought what are called “end-to end,” or more proverbially, “soup-to-nuts” solutions.  Indeed, after the announcement, a reporter got hold of a leaked memo from Cisco, an HP competitor that sells networking equipment, in which Cisco executives viewed Apotheker’s announcement with relish. The Cisco memo writer anticipates a big pickup in business, suggesting that, by “divesting the PC business, HP removes the bond between its other hardware, software and services groups, resulting in a disjointed company.”

Similarly, one trade website, Information Week Now, notes that after the PC decision was announced, “our polling of IT buyers showed that even though HP hasn't said a peep about printers, 32% of respondents said they were less likely to buy HP imaging technology now.” The website goes on to note, “On stage at Information Week's conference early last week, Lane, (HP’s board chair), acknowledged that 99% of HP's core infrastructure business was servers, storage, networking--the hardware that runs today's businesses. And yet everything HP has talked about for the past year--unstructured data, mobile, cloud--while important trends for the future, isn't where its customers look to HP today. It's fine to paint a vision of the future, but it's not fine to leave your customers behind.”

This last phrase, “leave your customers behind” is telling. If HP is to become an enterprise software company it will indeed leave some of their current customers behind and perhaps many of them in the lurch. In this sense HP’s customers are a drag on its evolution, and their current claims on HP’s resources could undermine HP’s future prospects. To be sure, by selling its PC division, HP would certainly lose a lot of revenue, but it would also accrue the capital it needed to purchase more profitable software businesses with their own new revenue streams. I suggest that Apotheker was fired because the board believed he could not manage this transition. 

This perspective sheds some new light on HP’s claim that Autonomy engaged in fraudulent accounting. The concept of market exchange highlights the idea of reciprocity. Money is offered for value proffered. But from a psychodynamic perspective exchange entails dependency and it is not uncommon for providers to resent the customers on whom they depend. For example, academics have an old saying, “if it were not for the students the teaching would be great.” These feelings highlight, that while exchange entails reciprocity, it may also mimic war or conflict. Customers always want more than the provider can profitably give. This is also why the phrase, “buyer beware” (in Latin, “caveat emptor”), has such an ancient lineage. 

This perspective also sheds some light on an interesting turn of phrase Meg Whitman used in describing HP’s strategy to its customers. “HP's strategy is quite simple; to provide the solutions for the new style of IT. We'd like to engage with you to provide the business outcomes you need and frankly require in the coming years." I find the use of the word “frankly” suggestive. As the term suggests, Whitman is telling her customers something they may not want to hear, an aggressive act, hence her need to be frank. And as she suggests, she can tell customers what they need, not the other way around.

The idea that one’s customers are one’s antagonists is to some degree a forbidden idea. So much business lingo is organized around the concept of serving customers and becoming their partners. Psychoanalysis suggests that while forbidden ideas may be repressed or displaced, they never disappear. This suggests that HP’s war against Autonomy- charging it with fraud, is a displacement of HP’s attack on its own customers. In effect, Autonomy is a scapegoat for the very difficult strategic challenge that HP faces. This is consistent with the idea that Autonomy was also once a stand-in, so to speak, for the new world of enterprise software, when it was first purchased. After all, looking at the purchase practically, it was Apotheker’s third choice. In both cases as stand-in and scapegoat, it was not just a company with real assets, but a symbol of HP’s prospects as well as difficulties. Indeed, it is common that people, who at one point are seen as heroes, are later seen as laughingstocks. As the song says, “the bigger they are, the harder they fall.” The deflation is the flip side of the inflation.

It may seem strange that a symbol can stimulate fundamental business decisions such as charging a company with fraud. But we have only to remember the passions that are stimulated when for example a citizen burns his country’s flag in protest.  Materially, the flag is simply piece of cloth, a rag, but symbolically it represents a panoply of thoughts and feelings. The ways in which we symbolize settings, objects and people shapes the concrete ways through which we relate to them. When Autonomy became a symbol of HP’s entry into the world of software it also became a symbol for the frustrations and difficulties of entering that world.

My prediction is that HP’s charge of fraud will not stand, but that its strategic challenges will persist.


  1. Among famous architects is the phrase, "My job is to subvert the client."

    We constantly speak in metaphors, and that use of language can hide what is really going on.

    I love my HP printer.

    1. Norbert. nice quote from the architects. In some ways they have an independent frame of reference- buildings as art-- which allows them to tell clients what they should be asking for. This is not the case with a commodity product such as a PC

      I am not sure that your reference to metaphor gives the whole picture. when HP called Autonomy a fraud ,they were, from their conscious point of view, not talking in metaphors. They really meant it. The idea that Autonomy stood for something else- namely customers, is held unconsciously and is specific to the situation. Looking at it generally, I think the reason why a metaphor is in fact a metaphor is because its reference is or can be widely understood. (e.g. "HP is running a race against time"). But the linguistic equivalence between HP's customers and Autonomy, connected by the feeling that both are enemies, is specific to HP's situation.

  2. Big decisions are attended by people who make things seem rational. They moderate psychodynamic influences, but I can see that, in HP's case, something slipped through.

    I'd emphasize that psychodynamics are already baked into the software market, prior to any acquisition. Autonomy’s technology is generally considered inferior. It was impressive when it first appeared, but it should have died. You can make good arguments for either high or low valuation. Sleazy accounting only compounds the uncertainty. Software is indeed the wave of the future, but the software might all be free open source with low-priced services.

    I am currently experiencing one of these leader flame-outs. The owner hired me to tidy things up for a CMMI rating which would unleash lucrative government contracts. He had a wildly elevated notion of what I could do. He bragged about having a Booz Allen/SAIC PhD on staff. In his mind, I lent a lot of legitimacy. He proceeded to overspend and is in deep trouble. After everybody took big cuts and worked hard to dig out of the crisis, he is now arguing that our work is bad and that we cheated him, and that therefore he is justified in not paying wages, though he still won’t tell anybody that they are fired. (He has an image of himself as never giving up on people.) I may not be paid for all of January.

    Why do people like this end up as the leader? Could it be that rational, no-drama people like me seek out a rain maker, and this is what we get?

    1. The idea that software stimulates on its own account psychic projections is very interesting. I would like to think about this some more

  3. It is interesting to note Autonomy as a symptom of dependency on existing and on prospect customers and on the board of directors.
    To me this emphasises the idea of reciprocity mentioned in the article, and the fantasy that life will be so much easier without customers. HP was looking to become more autonomous, and, instead, is becoming more dependent.
    I enjoyed reading this, thanks!