For service businesses of nearly every stripe, particularly in North America and Europe, manufacturing provides a glimpse into the future. Under unrelenting competitive pressure, they are reexamining the role of operations in creating competitive advantage and asking themselves which clearly differentiated services they can provide and how they can deliver those services to customers as efficiently as possible—and more effectively than their rivals do. Today's leading industrial companies asked and answered similar questions two decades ago, and many executives of service companies now believe that they can adopt the methodologies and tools that have already transformed manufacturers.

Insurance companies are in the thick of this transformation. To be sure, for many years their investments generated revenues that shored up their profits, so they experienced less pressure than manufacturers did to overhaul every aspect of their operations. But when the boom in equity markets cooled off, insurers had to earn their profits by selling products and serving customers—in other words, by running their businesses more effectively. A number of global insurance giants are thus overhauling their service and operating strategies. A few, including Paris-based AXA, are adopting tools and practices pioneered by manufacturers to help them improve the way they deliver value to customers.

Claude Brunet, the AXA management-board member responsible for operations, communications, and human resources, is in a unique position to discuss the lessons that service companies can learn from manufacturing: in April 2001, before joining the insurer, he was chairman and CEO of Ford France Automobiles, where he successfully implemented a number of improvement programs. In an interview with Eric Monnoyer, the leader of McKinsey's IT practice in France, and Stefan Spang, the leader of the IT practice in the United Kingdom, Brunet spoke about the risks and benefits of deploying industrial tools and practices in a service environment and about IT's role in helping to deploy them at AXA.

The Quarterly: What are the primary challenges facing a global insurance company today?

Claude Brunet: Most insurance companies are again striving to generate profits from operations. This effort has been a challenge for many of us. During the 1990s, the boom in equity markets gave insurers a healthy income from investments, which helped generate good profits. But when the boom ended, it became very difficult to generate them in this way. Many insurers were suddenly in trouble, so everyone went back to basics. The goal was to deliver profits from operations.

When you go back to basics, the first step is cutting costs. At AXA, for instance, we cut €1.3 billion in costs, across the board, during the past two years. Even in a difficult economic environment, we saw that we could generate significant profits. But improving efficiency is only half of the challenge; the other half is improving the effectiveness and quality of what we do. Why? AXA companies are differentiating themselves by the customer markets they target and the services they offer. To be successful, they will need to improve their operating performance along all dimensions constantly.

The second challenge is to generate higher organic growth. We'll do so primarily by introducing innovative products and marketing them in an innovative way.

The Quarterly: Let's explore both of these challenges and the linkages between them. What are you doing to improve AXA's operations?

Claude Brunet: We're using approaches that have worked very effectively at manufacturing companies and adapting them to our environment. These approaches include Six Sigma,¹ cost-modeling analysis, and internal benchmarking—benchmarking costs among AXA companies. Service companies can use these proven tools and methodologies to understand the weaknesses and strengths of their current processes and to focus on getting the results they expect. They don't have to invest time and resources to improve processes without understanding what the results should be.

These insights are very important for us. AXA companies must differentiate themselves—that will be their competitive advantage. The approaches we've adapted from manufacturing will help our people understand the processes of differentiation, how to improve execution by improving those processes, how to manage them for results, and so on.

We're at an early stage, so we can't declare victory yet, but the signs are very promising. We've made some very good progress in improving our procurement processes and our customer retention process. One country manager of the retention process was at first very skeptical, but he said, "OK, let's do a reengineering project using the Six Sigma approach, and we'll see." After the project was finished, he decided to become a Six Sigma black belt—a senior internal expert on Six Sigma improvement methodologies and tools.

The Quarterly: What challenges did you face taking an approach pioneered in industrial companies and transferring it to a service company?

Claude Brunet: The biggest challenge is getting service employees to understand that they use processes. After 30 years of Total Quality Management and other improvement programs, people in manufacturing already have this perspective. They know what a process is, how to analyze it, and how to improve it. This is all new for people in service companies. When we started implementing Six Sigma, early in 2002, we had to get our people to understand existing operations from a process perspective. And when you look at a service company's processes in detail, they may not be as repetitive as the processes and activities at a manufacturing plant.

On the other hand, I think that service companies accept Six Sigma more readily than manufacturing ones these days. Why? Because people in service companies have no preconceptions about Six Sigma. Employees of manufacturers have already struggled through a lot of improvement programs and are well educated in processes. You come to them with Six Sigma and they say, "OK, here comes another one." In service companies, you don't get that reaction, though you might hear people say, "This is a methodology from manufacturing, and we are a service company, so Six Sigma doesn't apply here." But you can overcome that kind of resistance more easily than you can overcome the resistance of people who have endured too many change programs. When employees don't know anything about Six Sigma, they are willing to try it even if they are skeptical. If you can get them to try it, they see its value, and—in this company, at least—if something has value, people will rush to use it.

So during the first phase of implementation you say, "Use this tool because it will help you to improve your process—and by the way, you don't yet know what a process is, but you will." We've gone through that. Now we're in the next phase. For instance, one of our companies started to work, very specifically, on differentiating itself through the services it offers. A detailed analysis showed this company that it has 60 processes linked to customer service. Now it will bring them under control in the next few years.

The Quarterly: What approach have you taken to implementing Six Sigma? Some organizations believe that they need to make it a key theme. Others are more pull oriented: they get people to try it and become familiar with it.

Claude Brunet: We've taken a sort of "pull-push" approach. We asked our ten companies to try Six Sigma, but we didn't want them to spend 20 months meditating on what they should do. We expected that as they saw its value, interest would grow, and that's exactly what happened. But you must structure the effort. There has to be corporate guidance to make sure that people go in the right direction. For instance, we initially set a minimum target of 5 to 6 black belts per country. Now that our companies are trying Six Sigma, we're raising those targets. In fact, we have about 300 Six Sigma black belts.

I think that a company's culture determines the approach you should take, in both manufacturing and services. We didn't use the same approach that GE used, for instance, because we're not the same kind of business. On the surface, the two companies look similar because they are decentralized. But AXA came together as a large global corporation, through acquisitions, only during the past ten years.

So we are not as mature a company as GE, which has operated in certain ways for a long time. Instead we are a teenaged global company—still largely an amalgamation of local companies. Our local managers know their business. They want to be independent and accept the idea of getting value from the corporate center only if it helps them create value day by day. GE is much more disciplined in terms of common processes, common practices. It has more maturity on the performance-management side too. As a result, it can not only set objectives and impose sanctions if they aren't met but also prescribe the tools that help employees to meet them. Of course, people at GE, like people at AXA, determine on their own how to meet their objectives. But here, we really believe in individual autonomy, accountability, and responsibility on the local level.

GE enshrined Six Sigma as a central theme—its culture made this an effective thing to do. But since our culture is different, so is our approach. We didn't make Six Sigma a central theme; we sold it slowly by demonstrating its value. At AXA, when you work in the corporate center you have to make sure that everything you do delivers value locally. And in everything you do, you have to use the right approach and behavior to be sure that if something has value locally, people will open their minds to it.

The Quarterly: What about the other operational-improvement levers—cost modeling and internal cost benchmarking?

Claude Brunet: In a manufacturing environment, if you are well organized, you know all your costs in detail. Although we are now trying to get additional data on them, service companies have yet to reach that level. The more refinement you bring to your cost analysis model, the more you find opportunities to raise productivity and cut costs. In our industry we have models for fixed and variable costs, but the granularity isn't terribly great. We have many ways to improve our productivity and our competitiveness.

The Quarterly: Are you moving to an activity-based costing approach?

Claude Brunet: Not yet. We're starting to look at the costs of our main processes and to benchmark those costs, internally, across our companies in order to compare them. Again, we help people to understand the value of this approach. Cutting the cost of IT, marketing, and so on—that's fine, but there is a limit. We need to become more refined in the way we optimize the performance of processes, and our people are open to the types of tools that could help us do so.

Service companies like ours can also benefit from other approaches. For instance, manufacturing companies have pioneered ways of examining how long it takes to develop and launch products and then of reducing their cycle time. That's important for service companies too. When you can reduce the time needed to launch products, you can launch more of them than your competitors do.

We have a lot of work; for instance, we need to refine the tools we use for our cost analysis. That means IT. Today, it's difficult to get information on costs and therefore difficult to know how you can cut them.

The Quarterly: What role does IT play in the operational improvements at AXA?

Claude Brunet: IT isn't separate from business operations. Right now, we're trying to reduce our spending on "lights-on" activities—maintenance, really—so we can shift more of the IT budget to investments. We don't assume that if they rise, so should total IT costs. We generate the funds for investment from cuts made elsewhere.

We must also become better at using the resources we have. Look at mainframes, for instance. Formerly, when we developed applications, we didn't pay attention to the hardware we were adding. But now we realize that if you look at the architecture of existing applications, you can significantly reduce the hardware they need and therefore your need for new machines.

Again, manufacturing companies have taken these approaches for years because they had no choice; they had to generate profits from operations in very competitive environments. We're now learning to do that.

The Quarterly: Are you outsourcing or offshoring operations to help manage IT costs?

Claude Brunet: We do some "captive" offshoring. AXA has a company in India—AXA Business Services—that we inherited in an acquisition, but we are offshoring very gradually and selectively. I know that other companies have different opinions about this issue, but IT is a core capability for us, and we therefore need to control our infrastructure ourselves. Also, we see big opportunities for improvement and have teams that could actually realize them, so we want to capture the full benefit for ourselves. We may, however, outsource IT or processes when the investments needed to keep them in-house can't be justified by the scale of the activity; for instance, we recently outsourced the back-office side of fund management in our asset-management company. The level of investment needed to manage these funds administratively is just too massive given their relatively small size, so outsourcing makes sense. But I don't see many areas where it does.

Anyway, we've dealt with our IT infrastructure, but now we have to deal with IT applications—a more difficult problem. In our environment, we can't say to everybody, "Let's have one enterprise-resource-planning system in all our offices throughout the world." We must take a more decentralized approach. Also, our decisions about applications have to reflect the operational consequences for our companies: what they need to differentiate themselves locally.

The Quarterly: The other challenge you face is organic growth. What's the linkage between that and profitable operations? It could be argued that all financial products are alike, in the sense that little emotion is attached to them, and therefore the quality of the operations supporting such products is an essential part of their value proposition.

Claude Brunet: In the property and casualty business there is little room for innovation. The same is true for some savings products and life insurance. If you ask customers to name their insurers, most of them won't know. So it is true, in a way, that companies in our industry have difficulty differentiating themselves or making customers believe that it makes sense to deal with one company rather than another. But there are ways to innovate—maybe not big, breakthrough innovations, but innovative approaches to products and to the way we market them.

Within that context, we are innovating at the margins in products and marketing. In France, for instance, we offer protection for people who don't drive their cars more than 8,000 kilometers² a year. We have a competitive protection product in France aimed specifically at young motorcycle drivers. We're also differentiating the marketing offer; for instance, in the United Kingdom we are now launching a protection product comprising several subproducts. The subproducts themselves aren't innovations, but the way we bundle and address them to the target segment is a marketing innovation. We also seek to innovate in the way we serve customers. So there is more than one kind of innovation.

If you want to innovate, you must always do so in a cost-effective, predictable way, and for that you must master your processes. Excellent manufacturers know how to do this. We will too.