Enron's fragmented business units spent money on technology like there was no tomorrow. And now there isn't.
In October 2000, 14 months before Enron filed for bankruptcy, CEO Kenneth Lay stood in the auditorium at his Houston headquarters and announced a new program. Called ClickAtHome, it allowed 15,000 employees to order a new PC from the company and take it home for free. The PC came with 24-hour tech support and broadband Internet connections for as little as $US15 per month.
Lay received a standing ovation. And why not? Wall Street was keeping Enron's stock around its high, $US80. The assetless trading model that Enron had pioneered with natural gas futures was rapidly being applied to other commodities such as bandwidth. Enron had reported $US30 billion in revenue in the first half of 2000-compared with $US40 billion in all of 1999. A giddy Fortune magazine was for the fifth straight year naming Enron America's Most Innovative Company, as well as the top company for Quality of Management, number two for Employee Talent and in the top 25 of Best Companies to Work for in America. Soon, it appeared, Enron would be the nation's seventh-largest corporation.
At the time, ClickAtHome looked like a relatively inexpensive thank-you from Lay to the hardworking entrepreneurs throughout his company.
But in retrospect, ClickAtHome may have been the pinnacle of Enron's wastefulness with IT spending. Enron had spent millions to install unused database tools, build massive trading platforms that some say never worked and pay outside firms to produce work that Enron could easily have done in-house. Employees who lauded ClickAtHome at the time now say it was crazy, because at the time executives knew that Enron's IT was out of control: the company had already started an enormous centralisation of IT to reduce the rampant redundancy.
At the same time, the business as a whole was spinning out of control. According to the 205-page Powers Report, written by Enron board member William Powers Jr and released on February 1 by a special committee formed by Enron's board after the company's collapse last US autumn, some of Enron's partnerships and the debt they hid violated ethical accounting canons. The report faulted top executives for approving the deals, and it showed that half a dozen Enron employees made millions of dollars in personal profit from these partnerships at the expense of the company.
"Rather than thinking in smart business terms, they resorted to this quasi-legal malfeasance,'' says Jill Feblowitz, service director for the energy industry at AMR Research in Boston.
While there is no clear and direct link between Enron's profligate IT spending and its accounting scandal - although both were fuelled by same decentralised, entrepreneurial culture - several experts, including Feblowitz, believe the situation in IT would eventually have caught up with the company.
"They were like kids in a candy store with technology," she says. "They bought a lot of technology, but they never bought a big, soup-to-nuts system. They worked with an awful lot of vendors, but they also built a lot of their own systems. If these [accounting irregularities] hadn't come up, the IT inefficiency might well have come up to bite Enron."
Enron IT was as cutting edge as it was byzantine. There were plenty of great tools, but there was precious little planning. Any given piece of its technology was likely to be the best in the world - as much as a year ahead of the curve. And it didn't necessarily integrate with systems that it needed to work with. Says Randy Wilson, a consultant at KPMG who worked on a team that logged 30,000 hours of work with Enron, "The systems sucked. They ran a $US100 billion company on Access and Excel."
"We all knew that Enron was out of control," says Linda Richardson, a former vice president of IT at Enron International who left the company in 1999. "But after you're in that culture for a long time, you begin to think you're the one who's wrong."
Value Through Chaos
Enron has always been decentralised. From its early days as a modest oil and gas company in the mid-1980s through its explosive growth to an energy giant that defined new ways of making money in the conservative energy business, the company had set up its business units as silos, or islands.
Data was organised that way too. Each of Enron's operating companies had its own vice president of IT, and that person was free to dream up his own IT architecture. The VPs would also budget, buy, develop and maintain their own systems.
"Decentralisation allowed us to be more responsive to our customer base," says Richardson. "In IT for international, we had to support people on the road all over the world. We understood what it took to install networks and international communications links."
When Richardson started at Enron in 1992, Richard Kinder was president, and he fostered this decentralised culture. Under Kinder's leadership, says Richardson, Enron appeared to "care more about its employees" than it did under Jeff Skilling, who became COO and president of Enron in 1996. Skilling, she says, added a brazenness to the culture.
"He brought in that whole culture of arrogance and attitude that We're Harvard MBAs and we can do anything'," says Richardson, who is now vice president of IT for Caminus, an energy software company in Houston. Skilling, according to several ex-Enron employees, pushed decentralisation to an extreme. He encouraged Enron's business units to compete rather than collaborate. Practically anyone with a new idea was free to spend money, steal talent from other business units, and as one IT staffer recalls, "just do it. Do it until you're told to stop."
According to former employees and consultants who worked with the company, the way Enron administered IT violated basic management tenets, from cost cutting to integration to having clear ideas of how data flows through the network. Sources say that culture of decentralisation and speed led to redundant spending and that different Enron businesses often bought the same technology without attempting to work with other business units to create bargaining power or to benefit from economies of scale. Ex-employees talk about "islands of data", but it was even more random than that. Without standards, the same data existed in different forms on different islands. Entire teams were dedicated to the task of deciphering the data.
"The core systems supporting the main revenue-generating activities were very disjointed," says KPMG's Wilson. "There were major disconnects from deal capture to risk management to logistics to accounting. They all worked from different data sources."
Making those different data sources agree when it came time to track sales, evaluate risk, monitor investments and report earnings was a Herculean task, according to Jeremy Merritt, an Enron veteran who had several different jobs at the company, eventually settling in with an SAP support team. "They had teams and teams of people who had to comb through the data and massage it so that it made sense," says Merritt. "There was a lot of magic, transforming apples into oranges and oranges into apples. Preparing annual reports was a joke."
Critics foresaw this as far back as 1999. They talked about "the risky strategy of decentralised IS management" that "threatens to wreak havoc on the company's computing infrastructure". The notion that the strategy could backfire was largely drowned out by a chorus of approval from a multitude of Enron believers.
"I'm not going to debate for a minute that decentralised IS has costs," said Chief Accounting Officer Richard Causey in 1999. "But it has some benefits too." And that was enough to appease most.
Chaos Without Value
To understand the chaos of Enron IT departments, visit the call centre. Merritt's SAP team at Enron decided to support applications other than SAP, and they branded themselves the Integrated Solutions Centre (ISC). Of the 220 ISC staffers, 50 were part of the customer care group. Of that group, about 10 worked the tech support call centre.
Merritt reports that the call centre fielded 87,000 internal support calls in 2001. Assuming a 50-week, 40-hour-per-week work year, this translates to about 43 calls every hour of the workday for the entire year, or one support about every 90 seconds. In July alone, there were 12,000 calls.
Why so many calls? First, there were few standards, never mind standard configurations. IT supported whatever the user happened to have. Also, IT project approval at Enron was all but guaranteed if you could fill out a standard online application form and file a report that demonstrated a 38 per cent ROI on the project. Tech support was constantly playing catch-up with whatever the business unit happened to decide to adopt.
The company's devotion to speed compounded the issues. At Enron, things happened in one-third the time that projects would take at other companies. One project to integrate sales information with wireless devices went from conception to deployment in 12 weeks, something the consultant who worked on the project said normally would take more than three months.
Donna Muniz was a Web designer for Enron Online who implemented Commodity Logic, a system that automated delivery of trading information to traders' workstations (yet another concept that Enron hoped to spin off into a revenue-generating business).
"I was always surprised by deadlines," Muniz says. "And in the back of my head, I was wondering how they could afford all this. They never said no'. If you needed anything, they let you get it so that you could move fast."
But ex-employees say this breakneck deployment of state-of-the-art technology was done with little regard for a management plan. "We'd get these modules going and then out of nowhere they would decide to do things differently," says Muniz, recalling work she did with integrator Valtech on the Web site Commodity Logic. "Nothing was planned, it was just done. We'd get all these horses galloping, and they'd have no reins."
Some IT projects worked, some failed. One particularly expensive failure came from the Data Technology Group at Enron Networks. In 2000, this group of 100 invested in a database monitoring tool called Quest Spotlight. The software was intended to measure database and server performance so the team could optimise its systems and stop buying new servers if old ones were available. According to Jose Lazo, who worked in the group, the software cost $US2 million and was installed quickly. But after its installation, it sat idle for eight months.
"Only about three people total even knew how to use it," says Lazo, who was hired straight from high school in 2000 and quickly moved up from hardware support to Enron's SQL database group. Asked why the team invested in Quest Spotlight, Lazo says: "We just bought the best of everything."
In October 2000, at about the time that CEO Kenneth Lay was receiving a standing ovation for the ClickAtHome program, Enron was in the midst of a massive SAP deployment that represented a fundamental shift in the company's structure.
Enron was beginning to centralise management of IT. It had begun with the SAP project, which had been planned as much as two years earlier. According to former vice president Richardson, the initial budget was $US80 million. Merritt reports that the SAP system eventually cost $US150 million and had an annual operating budget of $US20 million.
As part of the big SAP rollout, functions like accounting and compensation were being standardised and then consolidated so that Enron, under CIO Causey, could begin to see the forest for the trees.
According to several ex-employees, Skilling and Causey were the driving forces behind the about-face. (Both Skilling and Causey declined to comment for this story.) Another driver was Philippe Bibi, the company's CTO and an IT executive who enjoyed a certain star status at Enron.
Bibi was a hero for EnronOnline, the energy trading system that generated most of Enron's revenues and was hailed as a revolution for the commodities trading industry. (After Enron filed for bankruptcy, UBS Warburg bought EnronOnline and relaunched it in February as UBSWEnergy.com. Some believed Bibi's system changed the very definition of a commodity future by allowing any number of goods (bandwidth, for example) to be traded online. Bibi left Enron in spring 2001 and is now CIO at Putnam Investments in Boston. He declined to comment for this article.
Former Enronites could only speculate about why those executives embraced the countervailing philosophy after more than a decade of aggressive, premeditated decentralisation. The most widespread opinion is that the highest-level executives had no choice - they knew the situation in IT was worse than most people imagined. SAP team member Merritt believes the motivating problem was the reporting of financial data: putting together the quarterly and yearly reports was such a staggeringly difficult task that the executive team demanded centralisation to make sense of all the silos of data that had been erected.
By early 2001, Enron had deployed most of the major SAP modules, including human resources, financial information, asset management and a tax module, but the company could not rid itself of its decentralised culture overnight. For some time, the old freewheeling Enron coexisted in an uneasy partnership with the new Enron. The two systems occasionally clashed, as they did with the deployment of an SAP tool called the cross-application time sheet (CATS).
CATS was supposed to help employees track their project hours. Enron wanted this tool to be part of a Web front end to SAP's HR module, called eHROnline. But in typical old-Enron fashion, the time line for the project was compressed. CATS was hastily patched onto the portal, and it reportedly didn't work well. Anyone who worked on more than a few projects was powerless to track hours.
Merritt's SAP team took a look at the system because, as Merritt says, "This is the module that arguably touched the most people, and it was giving us a black eye." Merritt wrote up a formal spec that estimated fixing the CATS module at $US176,000. It was never fixed, he says, because the director who ought to have fought for the project was unable to get the ear of the people who were approving projects.
By the summer of 2001, the effects of centralisation could be seen in many parts of the company. Enron Networks had seized control of the IT support for two business units. And Enron Energy Services was consolidating eight separate billing systems it used into one centralised system - a massive effort dubbed Project Genesis.
The company was also creating standards for the first time. Causey, IT staffers say, was behind much of this. Eventually Enron Networks planned to disburse those standards as binding rules throughout the company.
One thing that the company standardised was the handheld devices for traders, starting, typically, with a luxury model - a $US1000 Compaq iPaq with a wireless network card.
In the database group, Lazo had also shifted his attention to standardisation and centralisation. He was trying to halt the proliferation of new databases at Enron by first consolidating them and then standardising their management plan.
It had been common practice, sources say, for workgroups at Enron to buy a server whenever they thought they needed one, even if there was server real estate available somewhere else. Lazo's job was to corral all of the databases splattered throughout Enron. Lazo would walk through buildings and look for them, or he'd call people on the phone and ask them about what hardware and software they used. "We'd have a team of 10 buy their own server and use it themselves," Lazo says. "A small accounting group turned up with a $US50,000 server. It was just sitting there. No one knew what it was for. We found databases that no one could identify or take ownership of."
Lazo says he finally felt that he would finally be able to impose some order on Enron's IT sprawl, but his efforts came too late. The goal of a centralised data centre was a 2002 project. And "before we could get to that point", Lazo says, "well, you know what happened".
Wisdom Too Late
Last year, just before Enron collapsed, there was some evidence that centralisation was working, at least in the information technology ranks.
"SAP brought a standard to the table, and everyone was using it for at least some of their transactions," Merritt says. "We were just heating up. We had standard call centre procedures and we were about to do the cool stuff, the data mining. We were right at the cusp of a total quality program. We had completely shed our baggage."
In three months, Merritt had managed to cut support calls by 25 per cent - from 12,000 to 9000. He achieved an unprecedented 93 per cent satisfaction rate with call centre customers under the new model. He reported these results with a newly standardised reporting system, which gave executives - quite possibly for the first time - a clear view of one of Enron's IT landscapes.
Merritt had even begun to think that Enron could brand the ISC's standard procedures and turn them into something Enron could sell to other companies. "We were on a roll," Merritt says. "We had a good thing going. And then, the rug was ripped out from under us."
The rug of course, was also ripped out from under Web designer Muniz, who, like most employees, was fired in December 2001. One month later, she finally got around to opening a box that had been sitting in her home office for months. It was a new PC with a large monitor, courtesy of the ClickAtHome program. vInside Enron: Traders Rule Ah, guys we have a loose IT cannon here . . .
Value through chaos was the unofficial motto of Enron's decentralised culture. Not surprisingly, the chaos was greatest where the value was highest. The most profitable business units were the most technologically out of control. This explains the lavish waste of Enron Capital and Trade (ECT), the most successful Enron subsidiary. ECT could do what it wanted because that was where the traders worked, and the traders were the ruling class.
"I was waiting for the day that they'd build a special elevator for traders so they could get to and from lunch more quickly," says Jeremy Merritt, an Enron IT veteran.
To work tech support for traders was an important job. And there were rules. A service-level agreement dictated that support calls from traders required an under-10-minute response time. Whatever technology the traders wanted, traders got. There were no standards.
In Enron's heyday, the average trader packed about $US20,000 of infrastructure on his desk in the legendary traders' room at Enron's Houston headquarters. Traders were using four different types of wireless devices at the same time, and IT was supporting them all. They indulged in high-end dual-processor workstations and $US2000 flat-screen panels. Some IT staffers report that traders who fitted 12 such screens on their desks, although the average seems to have been about three or four.
Workers Unite, Overnight
Jose Lazo was fired from his IT job at Enron on December 4, 2001. As CEO of wireless consultantcy Rexton Media, Anthony Huang had also been unceremoniously dumped by the company. Lazo called on Huang at about 8 o'clock that night, looking for work. Huang couldn't hire him but offered to help him set up an ex-Enron employee advocacy site. This is what happened next:
8pm: Lazo and Huang search for a domain name that won't present copyright issues. They buy "Enronx.org" - the .org to demonstrate their non-profit, worker solidarity tilt - for $US35 from Register.com.
8-9pm: Figuring they can't afford an ISP to host the Web site, Huang and Lazo call on a Texas IT advocacy group - called Techxans.org - to host the site. Techxans offers up a small bit of server space, and Lazo and Huang point the Enronx.org domain to that server's address.
9:36pm: Jose Lazo posts the first message at the Enronx.org Yahoo! discussion board, even before the site is finished. There would be 10 posts by 2am.
3-8am: Lazo, Huang and three others start filling Enronx.org with content. A company with some interface design in exchange for ad space. "Enronx.org: Discover the power of WHY NOT" reads the title bar with an X logo bastardised from Enron's E. The GUI is meant to poke fun at Enron. The conference room design is taken from an actual conference room at Enron headquarters. A picture of a gridiron player with a drooping gut is included. That picture of a fat, immovable object is meant to mock Enron's Web site, where a bicyclist, blurred by speed, is depicted.
8-8:30am: Lesley Plotkin, in public relations for Techxans.org, writes and distributes a press release hyping Enronx.org.
9am-noon: Lazo and Huang continue to fill the site with content, and they strike deals with companies to place ads on the site at $US300 a pop.
Noon: A local television station arrives to film a news story about Lazo, Huang and the new site.
The next day, 400 employers called to post jobs. Then, law firms thinking about class-action suits set up camp. Before long, there was more content than Lazo and Huang had imagined, including insurance information, basic workers' rights links and even a temporary grocery service through Kid Care, a non-profit that fights childhood hunger.
In the first 24 hours of Enronx.org's existence, there were 129 posts to the message board. By mid-February, there were more than 4000. About 4200 ex-Enron employees became members in the first five days. The site peaked at 6000 members in seven weeks. Currently, about 15 people opt-in per day, while five opt-out. And it's not just ex-Enron employees who become members of Enronx.org. There are many interested third parties joining or just watching the dialogue at the site's discussion board. Huang says the US attorney general's office has opted in as have countless members of the media. Even Senators' aides are members.
Lazo and Huang have added a newsletter, in which they highlight posts from the message board and link to important news articles. That newsletter is delivered for five cents per member. The newsletter also scored the duo their first big mistake: they had been highlighting comments from forum members who didn't want their comments highlighted. The media was picking up on this, and some members lost private information to the media.
Huang and Lazo have spent $US47,000 and have taken in $US23,000. Their next goal is to launch an effort called the Enronx Fund to garner financial assistance for those ex-employees in need.
Enronx.org demonstrates that despite our new cynicism about the Web, its speed and universality can be harnessed for those in dire straits or those looking for power in numbers. Workers can get more information, network with more potential allies and unite more quickly than ever using a Web portal - even one that was set up in one night.
"How long will it go?" Huang asks. "To be frank, I don't know. We thought it would go three months. It looks more like a year now. This is absolute grass roots, though. This is the American way done right. People get trashed by the bureaucracy, and now we can demand our rights. The money we spent, it's worth it."
Inside Enron: A Consultant's Dream
Enron's IT chaos didn't stop at the company's walls. It extended to third-party vendors. Sources report that even if Enron didn't need a consultant or integrator, empowered and ego-driven middle managers would routinely hire one. Web designer Donna Muniz recalls two incidents when her bosses hired outside companies and paid them hundreds of thousands of dollars to do work her group could have done.
On the Commodity Logic Web site design, Muniz recalls, "Out of nowhere, they hire this California firm [Organic] to do some design work. They fly them out three times. We promise to pay them $US100,000 to design the look of five pages." Muniz was put on the board that approved Organic's work. Privately, Muniz says, she fumed. She says that there she was, in meetings approving money for subpar work that her team was capable of doing.
"I asked [my project manager] lots of times," Muniz says, "Why are we doing this deal? I can do this work.' But my views never got past him. Eventually, I figured out Enron wanted to, I don't know, show off. They wanted to use firms that were popular. This design firm was hot at the time."
Anthony Huang, CEO of the wireless consultancy Rexton Media, handled eight wireless initiatives for Enron, mostly for the Enron Energy Services (EES) business unit. Huang characterises Enron's decentralisation and chaotic IT infrastructure as a consultant's dream. Imagine a world, he says, where consultants are paid to create incongruous and overlapping systems, knowing full well that eventually someone will have sort them out.
"We had a sandbox set up to pilot one wireless project, and we tried to make the service available to other business units," Huang says. "We'd say: Look how well this worked in Enron Energy Services.' But the other business units would set up their own sandboxes. I mean, it's the same sandbox, but they built their own."
"It went on and on," Huang says. "You'd end up with three different versions of the same software in three different business units."