June 2000 Something to Crowe About Is Level 3 CEO James Crowe Building the Next Great Communications Company? By Peter Horner There are probably easier things to do in life than build a global communications network from (literally) the ground up — making peace in the Middle East, creating cold fusion, beating Tiger Woods at Augusta National when he brings his "A" game — but what's the point? If you're looking for a real challenge, you go out and take on some of the biggest, most entrenched companies in the world, armed with little more than an opportunity, a vision, a business plan, and confidence in yourself and your lieutenants to make it all work. OK, it helps to have plenty of money, in this case several billion dollars, to wage the good fight. It also helps if you're well versed in the technical as well as the business side of the industry, you love what you're doing, and you have a great sense of humor to remind you and those around you to have a little fun along the way. Finally, it helps if you're packing a secret weapon called operations research, a weapon so powerful that it can help transform even fledgling companies into highly efficient, optimized organizations capable of competing on any business battleground in the world. James Q. Crowe brought all of that and more to the table in 1997 when he launched Level 3 Communications, the world's first end-to-end, fiber-optic, packet-switched, Internet Protocol (IP) network that has shaken the telecom industry to its underground copper wire roots. Crowe is counting on the cutting-edge technology of the "continually upgradeable" network to leapfrog over established players (who have massive investments in legacy technology to protect) and grab a significant slice of the global business-to-business communications market. If Crowe is correct, Level 3 will provide far faster transmission of data, voice and video than its bigger rivals at a fraction of the cost. Judging from the initial response of Wall Street and industry insiders, Crowe's game plan gives Level 3 something to crow about. A wildly successful high-yield bond offering raised $2 billion, at the time the largest single junk-debt issue ever. Another $2 billion in cash and assets came from Peter Kiewit Sons' Inc., an Omaha, Neb.-based firm where Crowe once presided over MFS Communications before teaming with Kiewit Chairman Walter Scott to launch Level 3. With $4 billion in hand, Level 3 hit the ground running in 1997 as one of the best-funded startups in history. A year later, before Level 3 had connected a single call, a confident-looking Crowe appeared on the cover of Forbes magazine under a headline that screamed: "Bell Buster: The big telcos wish James Crowe would just go away." In the article, Crowe said that he found monopolies "offensive" because they "stifle innovation" and "lead inevitably to waste." Crowe added that "introducing markets to monopolies is a lot of fun." Now, at the start of a new millennium, Crowe and Level 3 appear to be having a lot of fun, much of it at the expense of the traditional telecoms. With the five phases of its master plan pre-funded, Level 3 is well on its way to constructing a 24,000-mile network connecting more than 50 cities across the United States and another 21 cities in Europe and Asia. Level 3 has secured rights-of-way along 15,000 miles of U.S. railroad tracks to lay the network and has access to trans-Atlantic and trans-Pacific lines. As of April 2000, Level 3's network had reached 35 U.S. cities and five European markets, with a target date of 2003 for completion of the final phase of the original plan. By then, if all goes according to plan, Level 3 figures to be years ahead of the rest of the pack in terms of controlling a fully operational, optimized, upgradeable IP network. The man with the plan, the son of a World War II marine hero, wanted to be a mathematician or a physicist but claims he "wasn't smart enough." Instead, Crowe turned to engineering. He earned a degree in mechanical engineering at Rensselaer Polytechnic Institute and went to work designing power plants for Morrison Kundsen. In 1986, Crowe left Morrison to join rival Peter Kiewit Sons' at a time when deregulation was just beginning to alter the face of the electricity and telecommunications industries. Taking advantage of the opportunities afforded by deregulation, Crowe created MFS Communications in 1988 under the Kiewit umbrella. MFS became one of the first startup networks to compete successfully with a Baby Bell. After taking MFS public (to the tune of $1.4 billion) in 1993, Kiewit bought Uunet Technologies for $2 billion and eventually sold the group to WorldCom in a $14 billion deal. That was in 1996. The companies meshed but not the personalities. By 1997, Crowe was back at Kiewet, drawing up the business plan that would pave the way for Level 3 Communications. In mid-April, as Level 3 continued its quest to revolutionize the communications industry, OR/MS Today editor Peter Horner interviewed Crowe from Level 3's new headquarters outside Denver. Throughout the hour-long session, Crowe displayed his famous sense of humor, along with his legendary grasp of a rapidly changing industry that he and his company are primed to lead to the next "level." OR/MS Today: Are you building the next great telecom? James Crowe: Hopefully, we're building the next great communications company. Telecom has sort of a legacy connotation around here. We focus, for instance, much more on Internet protocol-based systems. From that point of view, it's a very different business than the traditional telephone business. But yeah, we think we are about the business of changing the communications market and the way services are delivered. OR/MS: Building a global communications network is a monumental task. What makes you think you can pull this off? JC: We were fortunate in that we started life as one of the best financed, if not the best financed, start-ups in history. We started with $4 billion in cash and equity, and we've increased that to $14 billion. We have about $8.5 billion of that available to pre-fund the basics of our global plan. There are very few companies that I know of that have the capital issues out of the way. What that means is, the rest is up to us — how well we plan and execute. That's an incredible advantage because it means that all of us, including me, can focus on planning and implementing, and not on where the next dollar of investment is going to come from. OR/MS: You've said that one of Level 3's greatest assets — along with the cash — is a blank piece of paper. What do you mean by that? JC: History has not been kind to traditional companies when there are step-function changes in the economics of an industry. When there are radical changes caused by, say, technology that drop the cost per unit for whatever product or service you're selling — the switch from mainframes to PC client servers, for example — the physics of the situation means that large corporations are going to struggle. It's hard to think of a single example where those same companies assumed a leadership role in the new market. I would say that, as a generality, the advantage lies with companies that have a good team of people, sufficient capital and a blank sheet of paper. The reason is, the older, traditional companies are stuck with legacy assets. They're stuck with their position in the capital markets where they have already worked their way through the start-up phase or the phase where you lose money and into a profitable position. It's hard for those companies to say to their stockholders and investors that we are now going to shrink or we're now going to start losing money, but trust us, in four or five years we'll be great. But most especially, they have legacy thinking. It's hard for a company that's been dominant to believe that they are, in fact, the underdogs and have to go through radical surgery. IBM, the icon of American business, almost went under before they turned it around. I'd much prefer to have those attributes I mentioned — money, a good team and a blank sheet of paper — and, by coincidence, we happen to have all three. OR/MS: Some industry experts and market-watchers say Level 3's real asset is James Crowe. Does that flatter you or frighten you? JC: It's a fact of life these days — and I don't think it's particularly healthy — that CEOs get far more blame than they deserve when an enterprise fails, and far more credit than they deserve when it succeeds. Part of it, as you know, is the media's desire to have a personality to write about rather than a faceless corporation. You can see the stark difference if you look at Microsoft and Intel. In a lot of people's minds, Bill Gates is Microsoft. Intel, meanwhile, has gone about its business and created an incredible amount of money, yet they get one-tenth the publicity as Microsoft because they are studious in avoiding the cult of the personality. It's silly to think that I caused all the problems or accomplished all of the successes here or anywhere else. Level 3 is a big company with a lot going on, much of which I don't even know about. All I know is that we have good people and I trust them to do their job. It's silly to think that I am what makes the business go, and it's silly to think that if I fell over dead tomorrow that another member of our strong team couldn't run the company. OR/MS: In a 1998 cover story shortly after Level 3's launch, Forbes magazine said the "big telcos wish James Crowe would just you would go away." The implication was clear: the big boys are either looking over their shoulder or already running scared. You haven't gone away. Should the big boys be scared? JC: If they were as scared as they should be — not just of Level 3, but of this revolution in communications — I'd be very nervous, because it's that kind of fear that results in the actions that are hard to take but are needed. I haven't seen any evidence of that, frankly. OR/MS: You once said that introducing markets and competition to monopolies is fun. Are you having fun yet? JC: If you promise not to tell our compensation committee, I would pay to do this job. Of the top 20 people at Level 3, 95 percent came from MFS. They left that company with sufficient personal wealth so that they would never have to work again for the rest of their lives. I'm fortunate to be in that category, but like all the rest, I'm here because I like it. OR/MS: So the answer is yes? JC: The answer is most certainly yes. OR/MS: MFS Communications was one of the first companies to compete successfully with the Baby Bells. Did you carry over any lessons learned from that experience to Level 3? JC: Absolutely. While MFS and Level 3 may appear to be similar companies with similar goals, they couldn't be more different. MFS was a company that created value based on a one-time change in technology — the shift from analog to digital communications and the shift from copper to fiber wire — and a one-time change in regulations — a shift from a monopolized market to a more open market. The changes aren't complete yet, but they're certainly heading that way. Level 3, at its heart, is a technology-based company and frankly, the central technology is operations research and optimization, which I tend to think is going to revolutionize all of business over the next 20 years. OR/MS: That's music to our readers' ears. JC: Operations research is in our business plan. I'm not saying it just because you happen to be from an operations research-based magazine. It's the heart of our strategy. OR/MS: You come from an engineering background. You started out as an engineer, and you're familiar with operations research and optimization. Has that background changed the way you, as a CEO, look at things compared to, say, someone who came up through the sales and marketing side? JC: Well, as much as I hate to say this after that nice introduction, I do actually have an MBA as well. The way the world is today and the way value is created really favors organizations that you could call either marketing technologists or technology marketeers. There are very, very few of them. If you want to personalize it, I'd say it's whatever Bill Gates is the best at: understanding the nexus of technology and what customers want. There are any number of examples of excellent technology companies that produce solutions to problems that don't exist. There are any number of marketing-based companies that know how to figure out what their customers want, but can't execute and get killed by companies with better technologies. The key is to have strengths in both. Hopefully, I have some amount of skill in those areas and I know I'm surrounded by a lot of people who are better at it than I am. OR/MS: Is it fair to say that you consider that engineering component of your background critical to your success as a CEO? JC: Frankly, I don't see how you can manage any kind of technology-based enterprise without it. Perhaps if you're Leonardo DaVinci, and genetically you understand technology, that statement isn't true. For the rest of us mortals, I don't see how you can understand the problem without a pretty good basis in technology. I don't think engineering is the only route. Applied mathematics is another. John Malone, who ran TCI for all those years, had his degree in OR. So there are lots of routes you can take to get there. But if you don't have a technology-based education, I think you need to be Leonardo DaVinci. OR/MS: I'm guessing you're a calculated risk taker. How do you look at risk in the business world? JC: In a mathematical way. OR/MS: Can you elaborate? JC: Risk, from our perspective, is the volatility around a predicted course. What I mean is, we view risk the same way Wall Street views risk. OR/MS: That plays right into the hands of operations research. JC: Absolutely. A good part of modern corporate America is built on strategic planning. At the heart of strategic planning is something called discounted cash flow analysis. That means figuring out what your cash-generating capability might be over the next five or 10 years by figuring out what your sales are and figuring out what your costs are. And then you recognize that a dollar of cash flow is worth more today than it will be worth in four or five years, so you assume some discount rate. That is, my money is worth 15 percent or 10 percent or 20 percent, and you discount those future flows at that rate per year. That's all strategic planning is today. That has revolutionized business. That's the way Wall Street views companies. Those analyses weren't really possible in any meaningful way before the computer. Even today, the projections about how much you're going to sell come from your sales organization. And then they're handed over a wall to your operations organization who figure out how much it's going to cost in all those years to produce those sales. In effect, you get a single estimate, when anyone would recognize that what is really necessary is to look at a wide range of possible outcomes and look at all of the possibilities, not just one. You look at one, you're assured of being wrong. Add to that the fact that we are in a very price-elastic business. By that I mean that for every 1 percent we drop price, we see a 2 to 3 percent increase in demand. That's no different than the market for microchips, software or routing on the Internet. There are lots of tech-based markets that have the potential to drop costs by 30, 40, 50, or, in our case, 70 to 80 percent a year. If you drop your costs and then you drop your price, you get a tremendous explosion of demand. Trying to model that means trying to model all the possible combinations of your network elements, your sales and services, and trying to understand what happens if I price my service at X, or one-half X or two times X. What effect does that have on demand? The more demand you have for something like our network, the lower your unit costs, and the faster you can drop the unit price. OR/MS: Again, that sounds a lot like OR. JC: You're right. When we first started Level 3, we literally toured the country looking for an organization we could work with to solve what is an incredibly large, nonlinear, integer optimization problem. We couldn't find anyone. The only people who really approached optimization or OR problems of this scale were the airlines, and they don't have the ability to drop their costs 60 or 70 percent a year. The problem would be similar if you could buy an airplane whose engines were twice as efficient every nine months, and that every nine months the same airplane could handle twice as many passengers. We finally hooked up with some operations research folks at Bell Labs. We worked with them on the basic algorithmic stuff, and in the interim, we built our own internal OR capability. Now, we largely do it internally, but we will be forever grateful to Bell Labs for helping us characterize the problem. OR/MS: So operations research was clearly one of the building blocks of Level 3. JC: Yes. Operations research was one of the major blocks in the foundation. Obviously, people are a major block in the foundation, and capital, but so is OR. Now tie all of that together. Being able to do a discounted cash flow model of your business has really changed Wall Street — the people who provide the money and the corporations who use it. Being able to model a whole corporation and then optimize those elements where you have tens of thousands of variables and tens of thousands of constraints I think is going to completely change business over the next 20 years on the scale that the Internet and the Web are changing business today. We are at the very early stages of that. The problems that we're trying to optimize in just one particular piece of our business involve so much computation and so many variables that even a short time ago — say three or four years ago — the tools didn't exist to do it. So we are at the exciting early stages of this when the payoffs to people who are able to apply these tools are enormous. OR/MS: Are your OR people brought in on important strategic decisions, or just the operational decisions that require optimization? JC: Well, we are learning about that. I'd like to say that they are fully integrated throughout the whole corporation, and that everybody appreciates the capabilities that operations research and optimization bring to our enterprise. But as you might guess, we have a lot of people who are focused on IP-based communications. The notion that they ought to be thinking about the 50,000 ways they can approach it as opposed to one and then looking for the best, that's not something they've heard before. A year ago when we started this, there was a fair amount of skepticism among a lot of people about OR. They could see the intellectual and theoretical potential, but there was skepticism about whether OR could actually deliver something that made a difference to the business in a reasonable amount of time. Today, there is widespread appreciation of the potential. Instead of me leading the charge and trying to push people into this, our optimization folks get lots of calls from other departments, far more than they can handle. A year from now I think that trend line is going to mean that we are fully integrated in regards to optimization. In fact, we have a whole team right now that is charged with building optimization skills into all of our business processes. OR/MS: In what ways do your OR people contribute on a day-to-day basis? JC: They're doing things like working with our engineers on optimal IP routing. There are other companies that do that. I think, though, that the current state-of-the-art in OR and the current state-of-the-art in IP rarely come together in industry. We do that. There are a lot of companies that go out and buy an optimization routine from somebody and then run their routing through it. There are very few companies that have a whole group of OR professionals who are trying to apply best practice and the best current thinking in OR to all of our operational issues. We're one of them. OR/MS: Clearly, you have a great deal of confidence in OR. Where does that confidence come from? JC: When I was in college, I wanted to be a physicist or an applied mathematician, but I wasn't smart enough. But I did learn enough to appreciate the power of operations research. Since then, I have spent a fair amount of time trying to stay current with the state-of-the-art and what's possible. I know there are things like interior-point algorithms. I know what they can do, but I don't have a clue how they do it. If you asked me to describe the interior-point method in detail, I'd have to call somebody in from our OR group. Let's say I have an appreciation for the power of the tool, and I try to keep myself aware of the tools and what they can do. I admit that to me there's a lot of a "black box" to it, but I try to keep up with it. OR/MS: Level 3 sounds like a great environment for an operations researcher to work. JC: We're talking to you because we want to convince everybody in your group [INFORMS] that reads your magazine that we are the place they want to be. We aren't talking to you to tip off all the other companies in the world about operations research. This is an exercise in human resources and recruiting. OR/MS: If that's the case, then I should warn you that this issue will be distributed to thousands of CEOs and managers in order to bring OR to the attention of corporate America. JC: I would expect you to, and I'm happy to be a part of it. I think it's going to be a while before a lot large corporations appreciate the power of what's going on. Remember, too, that when you say I'm going to bring operations research and, in particular, optimization skills into my corporation, you've got 150,000 people to deal with, you've got a number of older employees, you're probably heavily unionized. In a large corporation, it's a major thing just to get the thinking turned around. The intellectual property associated with integrating forecasting and optimization and using it to manage our whole business gives us a huge asset. It's the core of our competitive advantage, and it's not easily duplicated. OR/MS: You've made it impossible to overstate the importance of operations research to Level 3. JC: We gave a presentation on Wall Street that said precisely how we've integrated operations research into our business. There was a slide or two that described at a very high level what we've already discovered and how we've already benefited from the fusion of OR into our overall business. For instance, a traditional communications company would say, Well, we're going to go with a fiber-optic network. We'll put the fiber in the ground and that's the end of that. When we first started Level 3, we asked a lot of people the following question: What's the pace of change in optical fiber? And if you know the answer to that question, how do we build a network that can accommodate each new generation of fiber without opting to rebuild the whole network? In effect, what we asked was, If we are going to spend the money, how many empty conduits — which are just plastic pipes — should we put in the ground in addition to the first one which we'll use to start our business, so that each time a new generation of fiber comes along we can pull that fiber through the empty conduits at a cost of $100 million over six months versus having to rebuild the whole network, which is a multi-billion dollar undertaking in the United States alone? Nobody knew the answer, literally nobody knew. OR/MS: Yet another OR problem. JC: It sounds nice and simple, right? The techniques necessary to come up with the answer of how many conduits you need are at the very cutting edge of operations research. As I said earlier, it's a nonlinear, integer problem. It can't even be solved directly. You have to break it up into smaller problems. We said to Wall Street — and by the way, everyone in our business is now putting some number of conduits in — that we wouldn't give you the exact number of conduits because that's proprietary, but we will tell you it's in the range of six. That alone has changed the whole way in which the industry thinks about how they build their networks. That's a direct result of the operations research group in our company. Believe me, everybody appreciates them now. OR/MS: Was Wall Street impressed? JC: The average telephone company writes its assets off over 14 years. In essence, they say we have to rebuild our network every 14 years. We said, wait a minute. We have all this technology changing at various rates. Some things, like conduits, change very slowly, and it obviously won't cost you any less to dig a trench five years from now than it does today. Optical technology, on the other hand, is doubling in price performance every nine or 10 months, the most rapidly improving technology in industrial history. We said, we ought to build a computer model and figure out something that is absolutely essential to the financial management of any company: what's your average asset life and how often do you have to replace all of your assets? Guess what, the answer is something like four to seven years. Again, there has been wide recognition on Wall Street that that is a lot closer to the answer than 14 years. That impressed Wall Street and made a big difference in our successful offering. OR/MS: All of that analysis came from your OR people? JC: It did indeed. OR/MS: Here's another recruiting opportunity: Describe Level 3's corporate culture and how upper management responds to research and analysis aimed at decision-making. JC: We like to hire adult professionals, and my definition of an adult professional is someone who takes responsibility for keeping themselves at the forefront of whatever is their area of expertise. Then we support that kind of individual. Many companies have smart people to do analysis. These smart people come up with a conclusion, management finds that the conclusion does not fit well with their preconceived notions and decides that the conclusion must be wrong. What we try to say is, Look, operations research has the potential to revolutionize the world. If you can show us in a factual way, we'll promise to try to be smart enough to understand what you are saying. You make the point and we'll change our business. And we have. We can show lots of examples where that has happened. I'm not trying to pretend we don't have some amount of folks who say yada, yada, yada, but it's still theoretical. How does that help me today? But I'll tell you, there are fewer and fewer of those, and at the top of the organization there is a lot of push to make sure there are none of them in short order. I think also, just as a broad generality, there would be wide agreement that the single most exciting area today in any industry is networking. Period. That's where all the intellectual and financial capital is flowing. It's sort of where the desktop was 10 years ago — the hardware and software that the Microsofts and Intels developed. As you know, those companies recognized where the world is going, so they're all trying to turn into Internet-based companies. Well, we're from the start right in the middle of that, and we also tell everyone out there with the requisite OR skills that we think operations research is going to rapidly be the most exciting single area in business or the single biggest enabler in business. So if you're in OR and you want to be in an area where, if nothing else, you'll make yourself incredibly valuable to a whole lot of folks, give us a call. OR/MS: This is poster-boy stuff for OR. You don't find too many CEOs who appreciate operations research to this extent. If OR is everything you say it is — and you don't have to convince me — why aren't more companies using it? JC: Because only a short time ago the tools didn't exist in operations research to have the kind of leverage on a whole corporation we are talking about here. You'll find the best examples of OR-enabled organizations in the airline industry. In fact, we've spent some time with Sabre (see related story on page 46) because they wrote the book on the subject, and they couldn't figure out why rest of the communications industry hadn't embraced OR. The answer is, when you try to characterize or model something even at the start-up phase like Level 3, you couldn't get answers that were useful unless you shrunk the problem down to one little piece of the business. It is only in the last few years that the computational power and the algorithmic tools existed to see a way, a route, of characterizing a whole enterprise. That's powerful. We think our own OR organization would say we're not quite there yet. We've had to select a part of our business, a part of our network capability, and really focus on that. Otherwise, we end up with a problem with too many variables and too many constraints to solve. If you look at the rate OR is moving, however, I'm absolutely certain that in a few years you will be able to characterize and optimize across a number of variables that approach what you need to do a whole business. I don't know if that's three years or eight years away, but the trend line is clear. The payoff to those who get in early on the process is so big versus those that don't, it's on the scale of change from copper to optical fiber wire. We want to be leaders. OR/MS: Finally, put in perspective the impact that operations research will have on business going forward? JC: There was a time not that long ago when people were skeptical about the impact of the Internet and the Web on their business. Today, the conventional wisdom is that every company needs to be an e-commerce company, and the winners will be the ones that figure that out first. That statement applies directly to operations research. OR has the power to radically improve the efficiency of any company's operations. In the future, every company will be OR-enabled, and the winners and losers will be determined by who figures that out early and who misses the boat. We intend to be one of the winners. For further reading see the following sidebars: 'Packet-ed' and Ready to Fly Where Some See Risk, Others See Opportunities The Name Game Peter Horner (horner@lionhrtpub.com) is the editor of OR/MS Today. Table of Contents OR/MS Today Home Page OR/MS Today copyright © 2000 by the Institute for Operations Research and the Management Sciences. All rights reserved. Lionheart Publishing, Inc. 506 Roswell Street, Suite 220, Marietta, GA 30060, USA Phone: 770-431-0867 | Fax: 770-432-6969 E-mail: lpi@lionhrtpub.com URL: http://www.lionhrtpub.com Web Site © Copyright 1999, 2000 by Lionheart Publishing, Inc. All rights reserved.