GM's Ralph Szygenda has the Biggest Stick in IT

By CIOinsight  |  Posted 04-06-2006

GM's Ralph Szygenda has the Biggest Stick in IT

In 1996, when Ralph Szygenda took over the top IT spot at General Motors Corp., he was the company's very first CIO. His task: Take a highly decentralized company that had relied on one IT vendor, EDS Corp. (which GM owned) for more than a decade, and integrate its far-flung IT operations.

"It was an impossible job," Szygenda recalls. "EDS had been here for 12 years, and they were driving everything. It was assumed the autonomous businesses at GM would never come together."

It's now ten years later, and Szygenda is still CIO, as well as group vice president. He has cut the cord with EDS (mostly) and shaved billions off GM's annual IT budget. In the meantime, Szygenda has earned a reputation as one of the most influential CIOs in the world, and not only because of his $3 billion IT budget. He has acted as a visionary within the IT industry, driving innovation and standards by sheer willpower.

Now Szygenda is attempting to do something that no CIO has ever done: create a standard set of business processes among the most powerful, and fiercely competitive, IT vendors in the world. On Feb. 2, Szygenda announced the awarding of $15 billion worth of IT outsourcing contracts to the likes of IBM Corp., EDS, Hewlett-Packard Co., Capgemini and others. The goal: to create a truly global IT organization that could support the design, assembly and sale of cars and trucks anywhere in the world. Szygenda also revealed that the companies had been working together to create standards that could change the future of multi-sourcing. "We've got another difficult journey ahead," he says. "But I think we can do it."

CIO Insight Executive Editor Dan Briody spoke with Szygenda about bringing the vendor community together, how innovation is possible in an outsourced model, and how IT can help GM build better cars and trucks. An edited version of their conversation follows.

CIO INSIGHT: Give me a quick history of IT at GM.

Szygenda: In 1984, GM tried to integrate its technology better, so it bought EDS Corp. Ross Perot and EDS took tens of thousands of IT employees from GM, and they were put into EDS, but amazingly enough EDS supported GM the same way it existed before: as autonomous business units. So it created more profit-and-loss units than there were autonomous GM business units. This isolated the true GM company from information technology. There was one company running IT, and the rest was doing the other business stuff. It was kind of like two separate worlds.

That was what we call a first generation of outsourcing, from 1984 to 1996. GM went to one company, gave it the strategic responsibility for the technology, the driving of its business and everything else. Just basically outsourced it away. Costs went up drastically, and improvements were minimal, at best. But truthfully, EDS was owned by GM. So it wasn't all bad.

In 1996, GM split EDS off as an independent company. That's when I came in—the first CIO in the history of the company, which is quite amazing when you think about it.

GM was spending far more on information technology than any other automotive company, and we weren't changing the world. There was no efficiency. This was the second generation, from 1996 to 2003. We brought in all the major IT companies in the world to work at GM. And we went from 7,000 information systems down to 2,500.

So we cut a tremendous amount of costs from the environment. We spend $1 billion a year less on IT today than we did in 1996. I insourced 2,000 people. So GM took back the total strategic responsibility for IT, and the suppliers became only arms and legs to the company.

The third generation, which you're hearing about now, says, "Hey, GM has become totally global." We are running the business today so that we can design, deploy, manufacture and distribute from any part of the world to any other part. There are no regional boundaries at General Motors anymore.

So we are using those 2,000 people to integrate our IT suppliers, who were using dissimilar processes to run their own IT business. They all had different models. Everything from requirements definition to asset management. Not only did they have different models from one company to another, they had different models within their own companies, from one business transaction to another. That is the immaturity of the IT industry.

So you were able to change all that?

Yes, and let me tell you why. The IT industry started off on innovation, so each of these companies in their earlier days competed on who had the best product, and they didn't want to share. But they weren't really solving business problems in the companies they served. Their customers were left to integrate all the products.

IT vendors had very little knowledge of how companies ran, and they had very little knowledge of how to service them. If you look at Microsoft, they didn't have to worry about security issues. They just made a product.

I remember when I introduced Windows 2000, with Bill Gates, on the stage of the Moscone Center in San Francisco. I was the first corporate guy to work with Microsoft. They were talking about how they could do corporate work, but they didn't know what corporate work meant. They didn't do corporations.

The whole history of IT vendors says, "Don't work with each other." So you had a Larry Ellison and a Bill Gates, and the last thing they were going to do was work with each other. Maybe they even disliked each other. When people tried to integrate their products, it didn't work, and they got aggravated.

So people stopped buying those products. If you look at the growth of IT companies, it hasn't been that great, and their stock hasn't grown, because we're not buying innovation anymore. We still want innovation, but we want everything to work together, and until it all works together, we're not going to buy more. I'm serious: It is getting to be suicidal out there.

Given the complexity of the global environment, I knew that my model of 2,000 people trying to manage all our vendors every day would die in a few years. I had to change it, and the only way I could change it was to make it look like it's working as one, not many IT companies.

So two years ago I went to the CEOs of these major companies and said, "Hey, bets are off. This isn't ­going to work anymore." Now it does help that I buy $3 billion of IT every year. I probably have the biggest stick in the world. GM buys more IT than any other company in the world. So that helps.

I sat down with [IBM CEO] Sam Palmisano and [at that time H-P CEO] Carly Fiorina—and you can go down the list for every major IT company—and I said, "It isn't going to work in the future. The complexity's too great. You guys are causing all kinds of problems here. I've watched this industry for a long time, and you're really messing up. But realize you're causing yourself more problems than you're causing me. You're not growing. Go look at the stock price of Microsoft. Go look at Oracle's stock price. None of them grow."

And I said, "You've got to basically take the mundane stuff that doesn't differentiate and make it work together. If you don't have that in place, that foundation, you can't innovate any more. Even if you innovate, nobody wants it."

Next page: Making Changes, Not Threats

Making Changes, Not Threats


I didn't threaten them. This was two years before I was even bidding anything. I said, "Why don't all of you come to GM and for the next year and a half work together to help me design this. You are designing the processes that will help your business. By the way, IBM, you don't even do this across customers, you don't even do it across processes or business transactions. Neither does EDS. You can use this in your own company. I don't care if you share it with anybody else. As long as I have a two-year lead, you can take it any place you want."

I even took some GM money and paid for some of their people who were here. I actually paid them money to go define their standards for the future. Okay? And they worked in rooms at GM in secrecy, meaning they were not allowed to disclose this to the outside world, because as soon as it was disclosed, I'd spend more time talking with the press and talking with standards committees, and I don't have enough time in my life for standards committees. All I was concerned about was GM.

How did that effort go?

The first couple of weeks they were kind of strange bedfellows. They don't usually sit down in meetings like this. But it evolved over time. The CEOs wanted this solved because they saw the merit. You know, I'm not too sure they were totally enamored with GM.

I had people in each of the rooms who were the adult supervision. ­After about a month or two, though, they didn't need any adult supervision. They became a team, and they thought they were doing something that might change the industry for the future.

Now what they do with it, I don't know. Do I hope it affects the IT industry? Yes. All I asked of them was not to talk to the press. And for two years, it never got to the press, which was quite amazing.

I guess we're not doing our jobs.

Exactly right. You guys screwed up on this royally. Now why didn't they talk? They knew at the end there was $15 billion to be bid.

So it's a very interesting thing. I mean, if 17-year-old kids can collaborate over the Internet, build open-source software, basically run companies, design things, don't tell me that the best IT companies in the world can't collaborate.

So what is your role in this new scenario?

Well, I've always said that we are ­essentially IT brokers. All my people do in this new world is either build or buy information technology to change the business, and they are measured on improving the business, not the information technology. As I've told my people for years, if you go build a two-million-line program, and it's on time and within budget but it doesn't improve the business, you're the biggest failure I've ever seen.

Also, we architect the system. We don't ask our suppliers to solve our business problems. I mean, how in the world is EDS or IBM going to solve our business problems when they're not even in the automotive business? They can't.

So you're basically chief IT broker?

That's exactly right. I just manage a lot of money, but if you talk to my boss, I'd say 80 percent of my performance is based on business-process transformation, not IT.

Where does the innovation come from in this model?

I have what's called process information officers. They go horizontally across GM. I have a process information officer for manufacturing. I have a process information ­officer for product development.

I have one for sales and marketing. I also have CIOs in each of the businesses that have to make things work and drive the business.

The horizontals are the innovation. They are the economic buyers. They buy the new technology and the new capability. They are supposed to know what's in the marketplace, where it goes, and so forth. They are the ones who basically get involved with [venture capital firm] Kleiner Perkins.

My whole staff goes out to Kleiner Perkins ­every year and looks at the top 15 or 20 new start-up companies. These executives are the ones who buy the $3 billion a year. So they are the key brokers. If there are innovations they don't see, I'm going to give them all kinds of hell: "Where were you on it? Why did you miss that?"

Remember, I'm buying $15 billion, but we only talked about $7.5 billion worth of contracts. There's another $1.5 billion for telecommunications, but the rest of it—all the innovation—is being bought by the drink. So there's still more than $5 billion out there. It could be IBM, or it could be a 20-person company in Silicon Valley—or, these days, a company in India or China.

So how does this help you build better cars and trucks?

Well, what happens is it takes away all of that foundation building, and now I can start building the house. I can start differentiating.

Technology has a very, very significant position at GM. GM's processes run as well, or better, than just about every other automotive company. If you looked in the manufacturing area, three of the five most productive manufacturing facilities in North America are GM's—and that includes Toyota and Honda and everybody else. If you look at the product-development cycle time at GM, it is getting close to under 20 months, down from around 48 months. It's basically world-class. Those aren't issues at GM.

The issues at GM are pretty easy to explain but extremely difficult to fix. One, we have two and a half retirees to every active employee, and those retirees are getting full benefits and full pay. No other company has that ratio.

Toyota doesn't have to worry about pensions or healthcare. Other countries take care of that. But even at Ford, they have one and a half retirees to one active employee. GM has been successful for 100 years, and has a long legacy. That legacy will get fixed because our aging workforce will go away as they die. It's just going to happen, but we need another 10, 15 years.

The next thing is, because we've been around so long, our plants are in high-cost areas. So plants that were built in the 1950s and 1960s and 1970s are in the U.S. or western Europe or whatever, whereas our competitors put plants wherever in the world because they're just starting, or they're growing.

You can also talk about not having the right product. But that isn't driven by process so much as fashion. But you won't hear anybody talking about GM's processes being old and antiquated or not world-class. So if you're wondering why Wall Street is beating up GM? Well, they want to know what we are going to do with Delphi? What are we going to do with pensions? How about all this healthcare we have? GM's got to go fix those problems.

It has nothing to do with processes at GM and how the company runs.

So I think IT is a differentiator if we do it right, if we put the right model in. It can't change everything at GM, but it can help. All I'm trying to do is make GM somewhat successful.