Friday, February 16, 2007

Why businesses fail

A worthwhile analysis of the subject from :

Business failure is an opportunity for new CEOs to build fortunes and reputations. But with failure now happening at Internet speed, it is time to learn the lessons of history. Tom Hughes reports.

When in mid-September 2000, the owner of the domain name and Web site F**** (our asterisks) put it up for sale on auction site eBay, there were amazed reactions when bidding for the site reached $10 million.

The reactions were short lived. A day later the auction was revealed as a practical joke. Yet, in terms of Web metrics, the site is probably worth the money. It also has a dedicated user base that competes to pick dot-com failures and is awarded points for predicting disasters, from minor public relations errors, to buyouts and bankruptcies.

In the process, the company may have revealed the investment community in a harsh and unforgiving light, but the true value of F****dCompany is that it shows how businesses are collapsing for the same reasons they always have, but are failing faster than ever before.

In IT more than any other market, the beneficiary of an ailing business is often the manager who is parachuted in to turn around the company's fortunes. One such man is CEO of IBM Lou Gerstner, who is said to have reversed Big Blue's decline during his seven-year stewardship. For this, Gerstner received a $21 million signing bonus, has share options worth $90 million, and was paid a salary of $9 million in 1999. Many investors and disaffected IBM watchers, however, argue that IBM's turnaround is merely the result of cost-cutting and financial engineering - a very 1980s benchmark of management success.

A more timely example is turnaround king Charles Wang, chairman and ex-CEO of mainframe software company Computer Associates. In mid-1998, CA's board approved Wang a $655 million issue of stock if the share price of the company doubled. It did, although Wang (along with two other executives) was forced to pay back a quarter of the award in March this year.

Ironically, Wang's success was in founding and shaping a multi-billion dollar entity that preyed on failing software companies. But even he is not immune to failure. CA rushed out a profit warning on 4 July 2000 (Independence Day and a national holiday). Shares were hit hard the day after, falling 42.6% from $51.33 to $29.44. As a result Wang resigned, and took up the post of chairman, to be replaced by COO Sanjay Kumar.

The reason for the collapse was two-fold. CA had done something that the financial community dislikes: rushing information past it during a national holiday. But the profit warning, which noted falling sales of mainframe software, also suggested that CA's core business was failing faster than Wall Street expected, and called into the question the rationale behind its recent $8 billion acquisitions.

Kumar's CA now plans to sell off or list either its growth or non-core businesses, many of which - such as Sterling Software's Federal Systems business - it acquired just six months before. But the fact remains that the world's number three software company's September market capitalisation of $14.6 billion was roughly that of a hot start-up with a reasonable financial history.

Reasons for failure

For many managers, the idea of turning around an ailing company is seductive: they get the kudos of the company's reversal of fortune, and share options at a depressed price. But why do many companies fail?

Pat Martin is the new CEO of enterprise storage vendor Storage Technology (StorageTek), which has been in effective decline for years because of its inability to create significant growth outside its original high-end tape archive business.

Martin identifies the company's error: signing an exclusive three-year deal with IBM in 1996 to distribute its Iceberg RAID (redundant array of inexpensive disk) devices. But IBM - which was filling a hole in its portfolio - walked away from the agreement, launched competing products, and won a number of StorageTek accounts. The end of the agreement has hit StorageTek hard: this year its revenue will fall from $2.4 billion in 1999 to around $2.0 billion, mainly because of the shortfall into the Big Blue yonder.

Martin's solution is deceptively simple: "We have to grow," he says, echoing the familiar cry of any ailing corporation entertaining potential investors.

But far more interesting than StorageTek's insubstantial strategy, was its route to failure: it educated the market about a new technology, then failed to deliver it. In the process, it handed a $3 billion-a-year opportunity to EMC - which until the early 1990s was an unknown circuit board assembler.

Network attached storage pioneer Auspex Systems has similarly lost its way, admits new president Michael Worhach. The company made a bet on Windows NT in 1998 with its new low-end product line, only to see little demand in the market. At the same time, the engineering team lost focus on its core enterprise Unix products and, as a result, the company hit serious financial difficulties.

Auspex has now called in turnaround specialist Regent Pacific, which plans to put the vendor - whose competitor, Network Appliance, is increasing its revenue at 120% per quarter - back on track within a year.

Pleasing Wall Street?

"Nasdaq forces CEOs to be dishonest," claims Dr Zvi Marom, CEO of Israeli networking equipment manufacturer BATM Advanced Technologies. The reason is simple: companies' values are based on discounted cash flow models. This is the total earnings (as a measure of cashflow) that a company will generate over 20 years (normally), discounted for a return equivalent to base interest rates and a further reduction in the long-term risk.

So when a company experiences lower than expected growth, reports late, or indicates other underlying problems, its share price can take an unexpectedly large hit. And when it does more than one of these things - as CA did in July 2000 - the punishment is severe.

But the worst mistake in this area - which is often repeated in the IT industry - is to be caught cheating, as the share price collapse of mainframe tools company Compuware indicated after its own profit warning. The company had been riding a wave of acquisitions, but falsely booked future service revenue into its profit and loss account.

Even the successful players use creative financial techniques. Microsoft and Cisco have both been burnt in recent months by allegations that they in effect pay no income taxes, after reports that their share option plans allow them to claim tax relief against employees' share options. Such option plans are not shown on profit and loss accounts.

In 1999, the worldwide value of mergers and acquisitions rose by over a third to more than $3.4 trillion. Cisco, in particular, has used 'pooling of interest' mergers extensively - which the US Securities and Exchange Commission (SEC) is trying to block. The reason for the SEC's move is that such deals enable companies to make acquisitions look more financially attractive, as purchase accounting includes noting long-term charges to earnings for good will. The current glut of US mergers, therefore, is partly driven by the SEC's future plans.

Of course, there are good reasons for some mergers and acquisitions - consolidating market share, buying new growth businesses or acquiring new distribution channels. But the mainframe software business in the past few years has seen an unprecedented level of M&A activity as a result of companies' need to disguise their financial health.

Healthy companies' desire to plunge into the marriage game, however, introduces risks that may outweigh the potential returns. This should be a warning to CEOs such as Carly Fiorina of Hewlett-Packard, who is in the process of acquiring the IT consultancy arm of PricewaterhouseCoopers.

One report, ironically by KPMG, concludes that more than half of mergers and acquisitions destroy shareholder value, and a further third make no difference to the dominant party's overall health. Yet, in the last two years particularly, companies around the globe have jumped into bed with each other on an unprecedented scale.

Force-feeding the customer

Digital Equipment (DEC) was the IT success story of the 1970s. The company invented the minicomputer, and with a superb engineering-oriented culture was seen as a model company. But its reign did not last, because DEC failed to appreciate a change in the market: the emergence of Unix-based computers, which used the powerful operating system and the easy-to-use C programming language. Ironically, all of these had been developed on DEC PDP minicomputers.

The company brought out its own version of the operating system, VMS, but it was the equivalent of "bringing a pie to a cake-baking contest", says Dan Kuznetsky, vice president of system software for market research firm IDC, and a former DEC employee.

Combined with DEC's engineering-centric culture, imposed by founder and ex-CEO Ken Olsen, this often generated superior products, but at the cost of diversity and lateness.

"DEC constantly built wonderful products, but they weren't what the customers wanted," says Kuznetsky. A mistake often made by technology-driven companies - especially if they lack the will to shout against louder marketing voices.

The correct route for technology companies - whether in hardware and software and services - is to develop 'architectural control'. This is an idea first crystallised by two academics, Charles Ferguson and Charles Morris, in their 1993 book Computer Wars, which added to businesses' vocabulary the concepts of horizontal and vertical orientation, proprietary lock-in and architectural control.

In his 1999 book, High Stakes, No Prisoners, Ferguson takes aim at browser company Netscape for failing to do any of these things. Inexperienced lead developer Marc Andreessen also built a technological architecture that was later described by Java inventor James Gosling as "spaghetti-coded".

Worse, believes Ferguson, was the company's reluctance or inability to build any differentiation into its offering. This effectively meant that the company had no technical control over its product and, unusually for software, it got worse with every release.

So Netscape soon found itself in a position where it was losing market share, was making no profits and had a disloyal customer base. After spouting aggressive rhetoric at Microsoft for two years, the company was eventually taken over in November 1998 for $4.2 billion by ISP America Online, with the help of Sun Microsystems.

So even before Microsoft started aggressively bundling its Internet Explorer browser with its operating systems, Netscape had lost the battle, says Ferguson.

Believing the analysts

As bad as allowing engineers to develop bad products for a receptive market, is the reverse. Arguably more companies have come unstuck over hyping good technologies, but without first checking to see if there is a customer base.

Part of the problem is that some unscrupulous analysts will predict the existence of a market with little or no research to prove it. One of the best examples of this was Go, the developer of pen-based computing, which swallowed $75 million in venture funding but never created a marketable product. The company collapsed in 1992.

But the award for the most conspicuous failure to size a market correctly goes to US electronics giant Motorola, which spent $2.5 billion over 13 years investing in the global satellite phone service Iridium, only to see its entire investment disappear in late 1999, when the venture went bankrupt.

For entrenched companies not to attack new markets with new products is one of the primary causes of the slow death of companies, explains Harvard professor Clayton Christensen in his book, The Innovator's Dilemma. The problem is that companies fail to see the value of investing in disruptive new technologies which could destroy their existing high-profit margin businesses - or, in the case of more arrogant companies, are seen as too small a market to invest in.

The 'Ahab' factor

But while the innovator's dilemma is dangerous, management monomania is perhaps the most insidious and avoidable trap. The company that has shown damagingly obsessive behaviour has been network operating system company, Novell.

CEO and founder Ray Noorda, after failed takeover talks with Microsoft, became obsessed with the fact that Microsoft was trying to destroy his company - a focus that became so intense, ex-Microsoft CTO Nathan Myrvold dubbed him 'Captain Ahab' in 1993.

Even though Novell had successfully fought off Microsoft in its core network operating system business for five years, Noorda decided that he had to take direct aim at the industry's Moby Dick. He bought 20 companies, including Digital Research (an operating systems company), Unix System Laboratories and office suite developer WordPerfect (subsequently sold to equally mismanaged Corel) over a three-year period.

Even after Noorda retired in 1994, and his successor had divested most of his acquisitions, Novell was damaged beyond repair. In April 1998 Eric Schmidt of Sun was brought in to solve the problems. After a few years of Band-Aid strategising, Novell is again in trouble, and now looks as though it may never recover (see Company watch, CBR September). Novell fatally lost direction under Noorda, let its core products lapse and ceded market dominance. Since then it has suffered a steady decline.

Loss of focus

The lack of focus that wounded Novell is endemic in a number of other companies, and one of the key reasons is size. Companies become fatally weakened by their internal bureaucracies, whose slow decision-making leaves them vulnerable to new competitors.

This leads to massive conflict between divisions, some of which are being artificially held back by their older, less dynamic sisters. This is a problem specifically for the greatest IT company of them all, IBM, and calls into question the notion of long-term corporate recovery.

One veteran IBM watcher, Bob Djurdjevic, suggests that the company disguised static revenues throughout the 1980s with creative financing schemes, primarily by converting its mainframe leases to sales.

CA chairman Wang argues that that there is now a culture within IBM of booking revenues for long-term contracts earlier than they are due, creating a shortfall at the end of these agreements. This means that every five years or so, IBM will have to clean up its books and take charges to earnings for the revenue it has booked, but not received, believes Wang.

If true, this means that IBM can be expected to carry out one of these large-scale, exceptional write-downs soon. But with IBM's stock riding at an all-time high, this will be a hard thing for its management to even consider. With IBM delivering less than 8% revenue growth over the past decade, however, it is difficult to see how even self-publicist Lou Gerstner can classify this as a full recovery, even if the accounting has been scrupulous.

But IBM under less reactionary management could still reinvigorate itself. Other companies, however, are unlikely to have that chance. After the revaluation of Internet stocks since March 2000, the question is: do a huge number of companies have unsustainable business models?

If the Internet is genuinely the great leveller in terms of a start-up having the 'presence' online of a corporation, then start-ups have to consider adopting the marketing spend of a corporation. But using all available finance without generating earnings - 'cash burn' - has been the hallmark of the market correction.

Cash burn, however, is not limited to start-ups - even the one-stop shop of the Internet,, is battling for survival. The company has been forced into a position where it is going to have to start reducing losses, if not actually generating profits.

Anecdotal evidence from consumers suggests that discounts are falling from their old levels of around 40%, to roughly 10%. This tests whether consumers are sensitive to price, or are loyal to brand on the Internet.

More alarming is that many of the new business ideas on the Internet have significant holes in them.

For example, in business-to-business exchanges, predictions are already being made that the cost per transaction will fall alarmingly close to the cost of supply within five years. The upshot of this situation will be a brutal business where no one makes any money.

But if these new companies look back, they will see that the old computer industry has much to teach them: create new architectures, differentiate them, and protect
them. The final irony of business - or failure - at Internet speed, however, will be the possibility that Amazon's patent on 'one-click shopping', and its expertise in building ecommerce sites, become its most valuable assets.

No comments: