Can a discredited software vendor come back to life? Sure, if its software is unique.
Michael McGrath had planned to stay only a short while when he became chief executive of I2 Technologies in February 2005. He's still there, though the time hasn't been particularly fun. The Dallas software firm was a shambles. It had lost 99% of its market value since the tech peak and two-thirds of its sales volume, in part to software titans SAP and Oracle. The Securities & Exchange Commission filed suit against three of its former executives for falsely booking $360 million in revenue over four years. (Two cases have been settled, and one is ongoing.) Also, $315 million in debt was coming due, with only $285 million in cash on hand.
McGrath, who had been on I2's board since August 2004, said it was "like [when] you go by a burning house and say, 'I gotta do something.' I couldn't stand on the sidelines anymore. The problems were solvable, and I knew how to solve them."
Today I2 is still struggling to grow, but at least it's profitable, with its balance sheet in decent shape and its salesmen back on the streets selling the sophisticated production-planning software that the company's founders developed in the early 1990s. If McGrath can win back the respect of corporate software buyers, he will be one of the few ever to revive a fallen technology brand.
Until I2 came along in 1988, big manufacturers had scheduled their purchasing and production lines using crude spreadsheets backed up by basic inventory software. Sanjiv Sidhu and the late Kanna Sharma left jobs at Texas Instruments to start I2 as a test bed for using artificial intelligence to sort through hundreds of thousands of possible outcomes to solve complex production schemes. Their software could simulate manufacturing operations, incorporating constraints like factory capacity, availability of materials and space for inventory. Companies used I2 to find the optimal number of parts to buy, how long to run an assembly plant and what orders they could fill. Between 1996, the year I2 went public, and 2000, sales went from $100 million to $1.1 billion, with big customers such as Caterpillar and 3M.
Then came the tech crash. Oracle and SAP, previously content to resell I2's products, entered the supply-chain software field. Some big customers began to fume at I2 for overselling product features, especially how easy it was to customize the code. In February 2001 Nike publicly blamed I2 for causing it to miss quarterly earnings by one-third. I2's demand-planning software reportedly told Nike to make too many Air Garnetts and not enough Air Jordans, leading to $100 million in lost sales. Then, in 2003, I2 dropped a bombshell when it announced it would have to rewrite four years of P&L statements to retract false revenues.
"I2 grew so quickly it never had time to establish the management fundamentals," says McGrath, who had spent 28 years at the consulting firm he cofounded, Pittiglio Rabin Todd & McGrath. "We didn't even know where our cash was."
In 2004, I2 issued $100 million in preferred stock, convertible to an 18% stake in the company, to a fund run by Geoffrey Raynor, a distressed-tech investor in Fort Worth, Tex. Raynor helped bring McGrath in to shore things up. McGrath raised $43 million more by selling some businesses, reduced debt by $210 million and cut $100 million in annual costs by firing staff, closing offices and curtailing employee travel and use of cell phones and BlackBerrys. In one year I2's operating margin (in the sense of income before interest, taxes, depreciation and amortization) went from 13% to 30%. In 2005, I2 earned $43 million on sales of $337 million, its first profit in seven years. The income statement is a lot cleaner, as I2 now recognizes many sales only when each component of a contract is fulfilled, instead of booking all of it the moment a contract is signed.
I2 has to start growing again--not an easy task. Sales still appear to be falling, and the supply-chain software market is stagnant, expected to rise only 2.4% a year through 2010. But McGrath thinks he can wring more dollars from existing customers and attract new ones--at least those who haven't standardized with Oracle or SAP--by selling them new software that will tightly weave together I2's older applications with programs from other vendors, used to plan production, purchasing, sales automation and bid-tendering.
Currently these applications talk to one another only if a programmer sits down and writes the code that tells them how. That can cost millions of dollars. Little changes can ripple through every one of those program-to-program interfaces. So most companies simply don't do it, opting instead to make limited use of I2's programs.
I2's new product line comes with a library of standard software components that data-processing managers can plug in more easily. It allows I2 to build up a client's capability gradually, which pleases frugal buyers burned by massive investments years back.
ON Semiconductor likes I2's new software so far. A chipmaker with plants and suppliers in the U.S., Europe and Asia, on ships 300 million parts a week and has used I2's full suite of products since 2000, improving on-time delivery rates from 65% to 90%. Production simulations used to require weeks of tweaking one variable at a time. Using I2's new scenario planner, ON can spit out what-if results hourly, getting a view of its whole supply chain, says Charlotte Diener, a vice president at ON. The company plans to upgrade to I2's new demand manager this year.
VF Corp., the world's largest apparel company (which claims the title of the I2 customer with the most complex supply chain), is waiting to see how the new software works before committing. The company, which has 800,000 different products (counting sizes and styles) coming from 1,400 factories in Latin America and Asia, already spent the time and money to stitch its planning programs together manually, so it's in no rush to add another layer of software.
Ellen Martin, who runs the VF supply operation, says that if I2's new software cuts the customization needed in half, it will attract business.
"If I was a new customer, I think it would be a big selling point," she says.
At a recent $21, I2's shares are up 100% since McGrath took over, versus 18% for the S&P 500. But that's still only 12.5 times trailing earnings. Outside investor Geoffrey Raynor has sold 3.2 million common shares, but he still owns all the preferreds, convertible to common shares at $23. Raynor made a name for himself buying shares of Continental Airlines and XO Communications near their lows. Raynor is likely to make his third big killing here.
