Fallout over Autonomy revelations has implications for cloud players too

The fallout from the HP/Autonomy deal has sent shockwaves through the industry but there's really nothing new there

HP's decision to take a $5.5 billion hit on its $11.1 billion Autonomy acquisition has polarised opinion.

In the UK, some like TechMarket Review's Richard Holway have come to the defence of Mike Lynch, Autonomy's CEO, while others, started by raising a hue and cry but have since moderated their position, pointing the finger of blame firmly at HP. For myself, I take the 'caveat emptor' position.

At the time of the acquisition, Autonomy was attempting to transition to becoming a cloud player. It knew the days of flogging hardware and software were over but still had to keep its financial masters happy. It is the common lot of vendors used to selling up-front licences and the attendant maintenance fees which have made the software business such a magnet for those seeking riches but which makes selling software as a service so incredibly hard to make profitable.

In Autonomy's case though, H-P is claiming accounting irregularities. It should not surprise that Lynch comes back swinging, claiming that KPMG, HP's appointed investigators in the run-up to the deal found nothing and that Deloitte, Autonomy's auditors had similarly given it a clean bill of health. Who is right?

From what we know so far, Autonomy did exactly what it said and in accordance with accounting rules which are perfectly acceptable in the UK. US observers think that Autonomy's accounting would have contravened US GAAP rules. To be 100 per cent clear, no-one right now knows whether that is true or otherwise. All any of us have are claim and counter claim, along with isolated examples such as the TikIt and VMS deals noted in a Wall Street Journal article. Of the TikIt deal, it is said:

One transaction now under scrutiny by HP is a 2010 deal worth around £4 million ($6.4 million) that Autonomy struck with reseller Tikit Group TIK.LN 0.00% PLC, a British supplier of software to accountants and law firms, according to a person briefed on the deal. That year, Tikit's inventory skyrocketed to £4.1 million from £89,000 the prior year due to "a strengthening of our relationship with a key software provider," according to its annual report.

Autonomy recognized the total deal as revenue, the person said. But Tikit paid Autonomy only as it sold the software to clients, the person said ... Mr. Lynch said that under the accounting standards Autonomy followed, the company properly could recognize revenue on the books when all products were delivered to a reseller, as opposed to an end client.

See what I mean? The WSJ pads out its piece by providing titillating details of the CFO's underwear-buying habits, examples of demeaning behaviour among senior executives and Autonomy's aggressive sales practices and culture. If that shocks then you've not been around software that long. The kinds of thing the Journal 'reveals' are barely notable when compared to some of Oracle's routine marketing practices.

At the end of the day, what's the beef?

At one level there is no problem. Selling software under the old licence-and-maintenance model is so profitable that the occasional screw-up doesn't matter.

Time and again I have seen situations where software vendors throw money away on wasted projects, projects that could never deliver claimed value and flog licences for next to nothing in order to secure maintenance fees. Anyone who thinks that Autonomy's business practices are out of the ordinary in enterprise software should think again. They are the norm.

The kinds of practice our US cousins now say Autonomy conducted but which are prohibited under US regulation were the kind of thing that led CA into so much trouble back in the late 1990s. In the CA case, the company shifted the goalposts so many times it was almost impossible to understand under what revenue recognition rules the company was playing at any one time.

No-one is suggesting that Autonomy followed the same path although plenty are drawing attention to the so-called red flags they raised in times past. It is the classic 'I told you so' approach so common among those who shorted the stock but subsequently lost when an eager buyer came along to scoop up what it thought was the deal of the century.

The real problem as I see it comes in three forms.

First, our industry is poorly served by financial analysts who are often unashamedly trying to pimp stock valuations. I've been on the end of some of those conversations where problems inside a particular company are overlooked or trivialised for the sake of the bigger picture. A great example comes from the Journal's article where it at once says the detail of certain deals was questionable while giving a pass to those who didn't examine the accounts in enough detail to figure out what was happening:

While Autonomy booked the revenue from the VMS deal on its top line, the cost of the data it purchased was included in Autonomy's sales, marketing or other expenses—parts of the income statement that often aren't considered a key indicator of a growing tech company's core profitability, say the three people familiar with the matter.

I don't know what planet these 'three people' are on but it sounds familiar. Of course it matters. Sales don't appear out of thin air and anyone who looks at accounts on a regular basis should be questioning any major shifts on either the income statement or balance sheet.

The second problem is that very often, the explanations needed to fully understand what's going on require the mind of a forensic accountant. Those last two words alone are often enough to send most people to sleep. Who cares? The general answer is 'very few.' The dominant market mindset that growth on the top line is all that really matters clouds the judgment of many as those same people seek to justify stock valuations.

Finally, we have weak oversight and weak boards. When it comes to acquisition, I often see otherwise otherwise bright people's good sense fly out the window. It would not surprise me if that was the case with Autonomy. At the time, Leo Apotheker, HP's CEO believed the company needed to switch gears and make stronger software plays. It needed a bulky acquisition. Autonomy appeared to fit the bill. HP made a pre-emptive strike ensuring that no-one else would bid on the stock. At the time, many observers were sceptical about the deal but in the end, the HP board voted it through. And that includes Meg Whitman, the current CEO who is now yelling 'foul.'

Most recently I have been critical of SAP's SuccessFactors' deal. Despite SAP claiming the deal is a rip roaring success, I still think that deal will come back to haunt the company in some form or other.

From what I have seen, some of the practices SuccessFactors was conducting prior to the acquisition were in a similar class to some of the current revelations being thrown in Autonomy's direction. In the SuccessFactors case, the total revenue we're talking about is barely a rounding error when considered against the backdrop of SAP total revenue. You can almost make the same argument about Autonomy's revenue at the time of acquisition. The same goes for the treatment of commissions, a topic highlighted in reports about both companies.

I wonder whether the truth is more mundane. I suspect that HP paid an eye-watering premium for a company that it didn't really understand, it knew it and now finds itself having to take a hit as it tries once again to clean up its balance sheet act. Unless Meg Whitman plans on committing public management suicide what easier target than employees who are no longer there? In my analysis of the SAP/SuccessFactors deal I said:

As the old saying goes: marry in haste and repent at leisure.

The good news is that as we go forward, we may at last start to move beyond creative accounting and the use of Mickey Mouse measures such as EBITDA and 'non-GAAP income.' Instead, I hope we will concentrate on what is happening in the core, cash driven business models of the companies some of us attempt to observe.

If correct, then that means much finer scrutiny of the emerging cloud-based businesses where the business model is much less robust than that employed in times past.

To date, none of the major players such as Salesforce.com and NetSuite have proven particularly profitable when measured in conventional terms. And to date, almost no-one seems to care that much or even if they do, it is not enough to keep the sell side bulls at bay. I suspect that might well change.

Read more about: