When a company’s corporate core gets too far from its businesses, from the marketplace, and from its consumers, then a new organizational model may be needed. That was true of Kraft Foods when I returned as chief executive in June 2006. I had just spent three years running the Frito-Lay division of PepsiCo, where decision making was highly decentralized. That experience had reminded me how powerful it is when people come to work every day aligned with and focused only on the business, rather than on the internal organizational demands.
Kraft, too, had gone through periods of decentralized decision making, as I knew from my previous 22 years with the company. But the company in 2006 wasn’t operating that way. Not long after I arrived, a request to review and approve a pricing decision in the German coffee business arrived on my desk in Northfield, Ill. It created a lot of unnecessary work when people asked us in corporate headquarters to make decisions that were not our province. Moreover, we weren’t managing our brands and categories at the appropriate level to understand their competitive dynamics. We weren’t as nimble or responsive as we needed to be, and it was affecting our results.
Now, three years later, we’re delivering. We had an exceptionally strong year in 2008, on both the top line and the bottom line, despite the challenging macroeconomic environment. And there has been an equal impact on the effectiveness of our management. Today, I spend the bulk of my time on strategy, on understanding key drivers within our business units, and on putting the right people in critical positions — where they have much more responsibility than in the past.
One of the vehicles for achieving these results was the “Organizing for Growth” (OFG) initiative that we began at Kraft Foods in 2007. Rewiring our organization was one of the four key strategies of our three-year turnaround plan to restore the company to sustainable, long-term growth. To achieve this, we had to essentially dismantle the existing organizational matrix and replace it with a decentralized structure that gave our newly reorganized business units more direct lines of responsibility. This was an enormous undertaking, both in concept and in implementation. It involved changing reporting lines, structures, and operating units. And it was only one part of a larger change initiative that involved operational changes — building up our sales capabilities, reframing our food categories, and implementing new operating metrics and financial rewards for our executives and managers.
There is often a tendency to believe that when you make a structural change first, everything else will follow. However, structural, cultural, and operational changes are typically made together, and they influence one another.
This is the story of our reorganization and the changes that it helped deliver. It is an initiative that is best described by those within Kraft who participated in it. Their story — our story — follows.
Before the Turnaround
Karen May: When I came to Kraft in 2005, it was a company with iconic brands. Kraft had scale like nobody else. On an individual basis, the talent was amazing. Yet somehow, we were getting in our own way. We had functionalized to an extreme, and had lost focus on business results.
Rick Searer: How did Kraft ever get so centralized? To answer that, you have to understand how the company of today came to be. We are essentially a combination of businesses that have been bought and sold over the course of 25 years. Philip Morris bought General Foods in 1985, then bought Kraft in 1988, and put us together into a de facto holding company a year later. Oscar Mayer, Nabisco, and Jacobs Suchard in Europe all became part of Kraft through acquisitions.
Instead of letting the old Kraft Foods, the old General Foods, and the old Nabisco each have their own functions and operate as business units within a holding company, we moved in the early 2000s to a centralized model with very strong functional control, leaving the business units as not much more than marketing entities.
As a result, they were increasingly disempowered and disenfranchised. And increasingly, there was a slowed-down, thickening sense of process. Functional leaders were making decisions that might have made sense for the whole but clearly did not make sense for individual business units.
Dave Brearton: Up to a point, centralization was the right thing to do. If we hadn’t done it — if we hadn’t ripped out some of the costs and rationalized some of the infrastructure of the companies we bought — we would still today have a very disparate company with no real leverage of our scale. But the resulting problem, as we got into the middle part of this decade, was that we weren’t growing. Our earnings were going down. And we weren’t responding to changes in the environment quickly enough.
Brian Davison: The research and development function would have a goal for cost reduction. So would people in manufacturing and procurement. People were rewarded for their ability to hit their functional goals. They were good at that, but their linkage to the overall business performance and goals started to wane.
Mark Clouse: The operations group was incented for cost management and efficiency rates. Meanwhile, as a business manager — I was running the China operation when we started thinking about these changes; I run the Brazil operation now — I was chasing revenue. There was always this friction between the two sets of incentives, and the business unit leaders lacked the ability to holistically run the business. That was a constant source of problems — it got in the way of accountability and overall ownership of the business.
Lance Friedmann: An example from mid-2007 might show how siloed we were. The European chocolate business had developed a really cool reclosable package for their chocolate bars. You opened it up, took a bite of chocolate, and then put the rest back in and the adhesive resealed. It really helped freshness. They’d been developing this new packaging for some time.
Somebody was showing it at a strategic planning session in front of Irene, Sanjay, and the rest of the Kraft executive team (KET). [The KET, numbering nine people at the time, is composed of the most senior executives of the corporation.] And Gustavo Abelenda, who leads the Latin America operation, including a large chocolate business in Brazil, saw it and said, “That’s really cool.” Sanjay said, “You hadn’t seen that?” And Gustavo said, “No, I really hadn’t seen it.” A lot of lightbulbs went on. We suddenly saw how disconnected some of our teams could be.
This was before OFG — we had no mechanisms in place for making this sort of information flow. Now, we have something called category executive teams — multi-geographic groups responsible for sharing ideas in the areas of biscuits, chocolate, coffee, and powdered beverages. And, not surprisingly, today there’s a reclosable package in Brazil.
Mark Clouse: Another problem was how much time we spent, as business unit managers, communicating up the line. In a centrally run company, it stands to reason that the people in the center would need to understand your business at a level of detail that allows them to make good decisions. So we would prepare answers for whatever set of questions might be thrown our way. That took away from time we would otherwise spend developing our brands, building our businesses, and addressing the needs of the market or of our consumers.
Framing the Alternatives
Irene Rosenfeld: When I returned to Kraft in 2006 as CEO, I spent a lot of time talking to people about what was working and what wasn’t. I talked to employees and sought out customers. And I asked our board what issues they thought were getting in our way.
These conversations confirmed my sense that we had let the pendulum swing much too far toward centralization; we had a structure that was 80/20 global versus local. We needed to move that pendulum back. The only question was how. It was easier to look at possible alternatives because we all saw that what we had in place wasn’t working.
Gary Conte: Given where Irene wanted to go, it became very clear that we had to go back and create much more accountable business units (we now call them BUs), as a way of pushing decision making down the hierarchy, deploying resources where they were needed, and giving our leaders more freedom to act. Going to BUs meant raising the stakes for the managers we appointed.
The main idea behind OFG — shifting accountability and resource allocation into the business units — never changed once we started talking. There were some variations on the theme: We identified functions that would take a hybrid approach, with some control remaining in the center. And we always believed that some things needed to stay at the corporate level entirely — governance, for example. We just wanted to be sure that the things we defined in that way were very few and were of the highest priority.
At the same time that we wanted to get closer to the markets, we couldn’t forget that at Kraft Foods, big is beautiful. We needed to continue to find ways to use our scale to our advantage. So as we developed these accountable business units, we also gave a lot of thought to the collaborating mechanisms that would help us take advantage of our size. Our corporate functional experts are one such mechanism; these people work across our businesses to build capabilities and disseminate best practices. Another coordinating mechanism: the councils and networks we’ve established to pursue market opportunities that cut across business units or categories. Corporate strategic planning is a third mechanism, involving the identification of strategic platforms that business units must address in their own strategic plans. Health and wellness is an example of a strategic platform that started in corporate and is now important across our portfolio.
The Defining Moment
Irene Rosenfeld: We began talking about a sweeping decentralization in early 2007. I knew this would represent an enormous undertaking for the corporation. Had we been making this kind of change at a time when all was quiet on the business front, it would not have been as big a deal. But that wasn’t the case.
I lay awake many nights thinking, “Is this the right thing to do, and is this the right time to do it?” In the course of those sleepless nights, I came to the conclusion that we couldn’t get where we needed to go if we didn’t do it. How I felt was irrelevant, though — it was really about how the team felt, and whether they would get behind the decision.
I held many meetings with the various teams, talking about the pros and cons of a sweeping decentralization plan. The business managers supported it from Day One — no big surprise there; when all was said and done, such a change was going to give them more autonomy. As for the functional leaders, it’s probably fair to say that they were at different stages of acceptance.
In April 2007, I wrote a memo to the extended Kraft management team — about 170 people in all — announcing an initiative I called Rewiring Headquarters. (Later, we would change the name to Organizing for Growth.) I said that we would change our operating model to allow more decisions to be made by line-of-business managers, redeploy and eliminate some resources at headquarters, and make Kraft as a whole more effective. I promised we would come up with a set of specific recommendations by the summer. Suddenly we had a timetable.
The defining moment, as I like to think of it, came in July 2007, at a meeting of the Kraft executive team in our Tarrytown, N.Y., offices. This was the first time that I publicly declared to my staff: “We’re going to do this.”
Lance Friedmann: There was discussion and some comment. And then Irene very deliberately went around the room and said to the members of the KET, one by one, “What do you think? Can you support this?”
Irene Rosenfeld: My goal was to get us to a resolution in such a way that the executive team would own it and would be prepared to execute it, coming out of that meeting. But what I appreciated was that even the individuals who did not fully support the decision said so. It wasn’t just, “Let’s salute the flag.” It was more like, “I’ll do what the team decides, but let me tell you first how I really feel.”
Leading the Change
Gary Conte: With a corporate-wide transformation like this, it’s not easy. You’re challenging culture, you’re challenging behavior, you’re challenging self-interest. And there are a million details to iron out. What should the organizational structure be? Who should staff it? Which business processes should change? And what should the motivational processes be?
Within a month of the Tarrytown meeting, we had assigned teams to tackle the eight most important issues involved in the reorganization — what would remain at the corporate core; how shared services would work; people and motivators; the redesign of cross-organization collaboration processes; strategic planning and performance management; how the new model would play out in North America; how it would play out in our international operations; and communications. Each team had a work stream and a deliverable. And each team had at least two KET sponsors — that was critical. It ensured that the solution being developed was consistent with the new model.
The KET sponsors provided a necessary buffer between the team and the rest of the executives. Otherwise, you could imagine that when a team came up with a recommendation that significantly changed another Kraft executive’s role, there would be no one to run interference, validate that it was consistent with the objectives, and help facilitate the discussion. Finally, and perhaps most importantly, being involved in the change process drew the individual KET members into the program. They had some skin in the game.
Brian Davison: One fundamental question, once we had decided to go with accountable business units, was what were the right BUs? For instance, by the time of the Tarrytown meeting, our Canadian business had basically been taken apart, and most of the business decisions about Canada were being made in the United States. With OFG, we decided it made more sense to treat Canada as its own BU with its own dedicated resources, which is what we had historically done. We did a number of such realignments, with the goal of making decisions closer to the consumer.
Irene Rosenfeld: At the same time, we were conscious of the need to preserve our scale advantages. That involved thinking through what this new organization meant function by function, geography by geography. It was never one-size-fits-all. The model we implemented in North America looked quite different from what we implemented in Europe or in developing markets.
Sanjay Khosla: There was a good reason for our varied models. Kraft Foods International, which I was brought in to run, shared North America’s problem of being too internally focused; and as in North America, it made tremendous sense for us to push decision making down to the unit level — in this case, a region or country. Historically, what was good for one market was seen as being good for every market. And that clearly was not true.
Irene Rosenfeld: It’s vitally important, when you’re making a change like this, to have the right people in place. In our case, the people running the businesses could no longer principally be marketers; they needed to become general managers. And the functional executives had to move to being functional leaders, capability builders, managers of careers, and sharers of best practices. Not everybody was able to make that transition.
Karen May: In retrospect, it was the business leaders who thought, “What’s the big deal? We’ve been operating like this for a long time” who had the most trouble growing into their responsibilities. The stronger BU managers knew they were taking on a huge shift. They would say, “This means my team has to be different and my talent needs to be different.” When you heard that, you had more of a feeling of comfort.
Irene Rosenfeld: To give BUs a better chance of succeeding, we changed the way we aggregated data on the individual businesses and evaluated their performance. We had always been a very metrics-oriented organization, but by now we had binders and binders of data, and it had become impossible for any one person to understand what they needed to do. And we often had situations where the metrics for the functions were in conflict with the business units’ metrics.
Brian Davison: We decided to distill what was important to any given business’s performance, and present it in a scorecard that could be understood at a glance. We used green highlighting to show areas where a business was performing as expected, yellow to pinpoint emerging problems, and red to flag areas requiring immediate attention. The three main measures on the scorecard — which we picked for their alignment with shareholder value — were organic revenue growth, operating income growth, and cash flow.
I think most people would agree the scorecards have been a big change for the better. Before, we had a culture where, as a business unit manager, you’d want to prepare yourself to be able to answer any question. Now, to a great extent, you know what’s going to come up in your monthly performance meetings.
Gary Conte: We also took a fresh look at decision rights. We had a lot of policies, either created over the years to address internal needs or inherited from companies we acquired. There were policies for every aspect of global company-wide activity: accounting procedures, employment, acquisitions, materials, selling to consumers, business processes, cash flow, employee gifts, and so on. Some of the policies no longer made sense; others required too many levels of approval. We eliminated many old policies, updated others, and put all of them in a consistent format; we made them easy to search for on Kraft’s intranet. People can use this intranet site to ask for notifications of changes in the policies, or to ask questions about them, or to request an exception or suggest a change. Most importantly, if a policy isn’t on the site, then that policy sunsets. We no longer have people following policies that aren’t current or relevant. And the quick access that managers have to these policies lets them know the boundaries and the reasons behind them; it frees them to innovate without feeling they need to seek approval from some other person or place.
Irene Rosenfeld: We changed Kraft’s incentive structures in a way that linked unit managers’ compensation more directly to their individual performance. For example, in the past we might have incented the general manager of a country — China, say — based on the total performance of Kraft worldwide. Now, bonuses are calculated by weighing the performance of their individual businesses (for example, China) at 70 percent and the next higher level of aggregation (for example, Asia/Pacific) at 30 percent. That is just enough to encourage the managers to support the greater good, not just their own individual performance. We’re hopeful this will encourage managers to make the trade-offs that help both their units and Kraft as a whole.
Learning New Roles
Dave Brearton: For most functional leaders, I’d say, the reorganization has been a huge shift. Previously, those of us in the KET had big portfolios. You might have had responsibility for a large number of factories, or for the customer service group worldwide, or in marketing for consumer insights worldwide. Suddenly you’re influencing decisions, providing tool kits, or managing talent, but you’re not actually running a day-to-day operation.
Karen May: There’s an art to letting go, which is necessitated by the new structure. In their new role, these executives are like orchestra conductors. They’re not playing the instruments and they’re not writing the music. But they are making sure it all comes together. It’s not a matter of abdicating responsibility. The responsibility is huge.
Jean Spence: Whenever you go through a change like this, you do feel a certain loss. I used to have 14 direct reports, and I would spend time with each of them on project updates. Moreover, I was involved in approving the year-end performance reviews of their direct reports — about 100 R&D staff people altogether.
That’s now done on a local basis. The head of R&D in the cheese division worries about cheese. My job now is to think about the strategic projects — for instance, what’s the 10-year trend going to be, what are the implications for R&D, and how do we build that into our strategy. Before, to be perfectly honest, I wasn’t able to spend much time on these questions. I’m finding the new challenge pretty interesting.
Rick Searer: It didn’t make sense to decentralize every function — and we didn’t. The model needed to be more flexible than that. We knew from long experience that there were functions, in North America in particular, that were better off centralized. We ran information technology and human resources centrally because they were more efficient that way. We kept sales central for reasons of effectiveness — our wall-to-wall sales force provides the best in-store coverage for our iconic brands.
Where does it make sense to enable each business unit to be nimble, and where does it make sense to take advantage of the scale of Kraft? Our answer to that question informs the model we have today. And the answer is still evolving.
Mark Clouse: As a business unit leader, I find that my ability to engage my organization has been aided greatly by the fact that we are setting more of the direction. And as a company, we’re doing smarter things. We’re not launching products just because someone thinks a platform that sells well in one place will sell well in another.
These problems are now gone, and we’re focusing on the things we should be focusing on: the top line and the bottom line.
Nick Meriggioli: I can offer a specific example of how things have changed for the better. Toward the end of 2008, operating income at Oscar Mayer, the North American unit I run, was a little behind plan. In the past, the sales planning team might not have had much reason to try to fill this gap. Their efforts, and much of their incentive compensation, were based on driving top-line growth.
This time, they devised a plan to focus on higher-margin products and trim trade spending. That helped us deliver the number we were looking for on the bottom line.
The new organization and the mind-set within business units of “We’re all in this together” — sales and marketing and operations and R&D, all pulling in the same direction — facilitated this tactical adjustment.
Sanjay Khosla: Fundamentally, the role of corporate executives in this new structure has been to put the right people in the right jobs, and then get out of their way. And I’d argue that by going to the business unit structure, Kraft has done one of the most profound things in its history. We’ve unleashed the potential of our people.
The Challenges of Change
Irene Rosenfeld: When you make a change of this magnitude, it can take a while for people to figure out what to do. Some embrace that responsibility faster than others. At the same time, some of the functions have actually become, in effect, too hands-off.
Dave Brearton: The center still has people who can be parachuted in to help with issues. I think we have been a bit slow to react from the center in a couple of cases, and the BUs aren’t always anxious to get our help. It’s a delicate balance, knowing when to intervene, but we’re getting there.
Jean Spence: For me, and I think for a lot of us, the biggest challenge in decentralizing is making sure we get the benefit of collaboration across the enterprise. In the past — when the top R&D people in the business units reported to me and I had to answer to Kraft’s needs — if we created a promising new cookie technology in East Hanover, N.J., I could have the VP go over and share it with our Asia/Pacific operation. His day job may have been local in New Jersey, but a very important part of his role was sharing his innovations globally.
We’ve had to make sure to reinforce that ethic even though we have a less centralized reporting structure. We have several mechanisms in place. The highest-profile example is our category executive teams. We also have collaborative networks through which we share best practices.
Nick Meriggioli: Along with accountability, people need information. Our information systems were built to provide information to the business units, but not to the categories that make up those units. The people who oversee the hot dog, cold cuts, and bacon businesses — subcategories within Oscar Mayer — need to get data on their monthly results at the same time that I get the broader business unit results. With our new implementation of SAP, we’re making progress here.
Karen May: Another thing we are still working out is how to adopt more of the shared-services model. The idea is that each function defines its services and costs; the business units, within guardrails, get to choose the level of service they need or, in the extreme, to opt in or out completely. In previous years, the business units had no such flexibility — they would just get a huge cost allocation for shared services that they didn’t understand, and that they saw as a big dump into their P&L. There wasn’t a mechanism for them to exercise any control over it.
The process for calculating costs using shared services is still not as clear as we’d like it to be. Indeed, it’s fair to say the shared-services model is still in its infancy at Kraft — each function has done it in its own way. But the problems in the broader economy should give shared services a push forward, because of the method’s potential to help us control costs.
Jean Spence: Before OFG, the business units were able to say, “I missed my plan, but you left me nothing else to pull from my P&L besides marketing, and of course that’s going to make my market share go down.” They can’t say that now, because they make their own P&L decisions, and I think we’re getting more effort from them. We’re certainly getting better results.
In addition, we’re being forced to allocate costs more accurately. When we peeled out everything that people do — the functions had to list their services and what they really cost — it became clear that certain business units were getting charged disproportionately. That meant that some of our more profitable business units were subsidizing less profitable ones. I personally think the transparency will make us a better company, because now, if we can’t get the margin up in a particular area, it will force us into some tough discussions and decisions.
Dave Brearton: In 2008, in a further evolution of OFG, we gave our business units an internal cash-flow target and, with all the P&L and balance sheet levers they needed now at their disposal, they exceeded it handsomely. You’re not going to improve your cash flow unless the entire business works together. If the dots connect only at the office of the CEO, it isn’t going to work.
Rick Searer: Overall, I would say that OFG is working very well. The organization is adopting good practices and has embraced it very positively. I think it’s fair to say, though, that the flywheel is only just beginning to turn. That’s not too surprising given the dramatic change and the fact that we had a business that was in need of fixing as we entered 2007. So we are nowhere near the end state on this. One is never truly in an end state, anyway.
Irene Rosenfeld: When I talked to investors about it, they were often concerned. “You’re making this tremendous organizational change!” they’d say. But I knew in my heart that this was going to be a big enabler for us. It is a fundamental human desire to be in control of your destiny and to make the necessary decisions that will affect your performance. When we tapped into that desire, it took us to a whole new level. We’ve just about turned the motor on; now, we’re ready to make it hum.
Six Keys to a Successful Reorganization
The strategy+business team asked executives at Kraft Foods Inc. what advice they would offer a CEO or senior leader considering a similar corporate transformation. The following factors came up most consistently in their answers.
1. Start with the business strategy. Whatever else it may be intended to do, the new organizational model should primarily enable and catalyze the strategic direction of the company. If the strategy isn’t clear, the organization cannot align behind it. In Kraft’s case, the need to move decision making closer to consumers and markets was a clear strategic shift that prompted other changes and discussions.
2. Go beyond lines and boxes in designing the organization. You must have the right people, and the right reporting relationships, in any change initiative. However, that alone is not sufficient for success. Kraft went further, addressing work-flow processes, decision rights, metrics, career paths, corporate policies, incentives, and talent development. Executives credit this deeper execution for many of the benefits their change initiative has produced.
3. Understand that one size does not fit all. Any robust organizational model will apply in different ways in different situations. Decentralization of the business units was Kraft’s theme. But there were also functions, and markets, where centralization served customers — and the company — better, because Kraft was able to leverage its scale. Where this was true, as in its North American sales operation, Kraft didn’t hesitate to make exceptions.
4. Have thorough planning discussions pre-launch, to be able to move quickly later. Major change initiatives work best when key stakeholders have had a chance to articulate their concerns — and when the team involved in the decision has had the time to think through the implications of the new model and establish some guiding principles. The detailed quality of these discussions can engender some initial frustration, but at Kraft they built a sense of collective ownership. The discussions also led to a road map that allowed the corporation, in the long run, to move more decisively.
5. Leverage the power of leaders. Transformation efforts that don’t have the support of senior leadership often fail in spectacular fashion. But Kraft required much more of its leaders than verbal support, asking its nine-person executive team, and then an expanded group of 120 senior managers, to hammer out the details of the new model and keep it moving forward. Kraft executives were involved in all major proposed changes — critiquing the changes, keeping them on track, and talking them through with other Kraft executives who might be affected.
6. Expect a multiyear journey. Even the best-planned change initiative requires some course corrections — or ends up working best when done in phases. Kraft began by moving decision making about such matters as product development and manufacturing to lower levels in the hierarchy — thus taking aim at one of the biggest perceived problems in the company, its general managers’ lack of control over key P&L levers. In 2009, Kraft is focusing on some other aspects of the new model, such as how to implement shared services. And now that the new organization structure is in place, the company is shifting its attention from rewiring the organization to building high-performing teams.
Reprint No. 09307
- Irene Rosenfeld is chairman and chief executive officer of Kraft Foods Inc. She has spent most of her career at the company. From 2003 to 2006, she was chief executive of PepsiCo’s Frito-Lay unit.
- Dave Brearton is executive vice president of operations and business services at Kraft Foods and was one of two executive sponsors of the company’s reorganization effort.
- Mark Clouse is vice president and managing director of Kraft Foods Brazil.
- Gary Conte, vice president of human resources, corporate functions, and employee services, served as one of the program leaders on Kraft Foods’ reorganization.
- Brian Davison, vice president of strategic planning at Kraft Foods, was co–program leader, with Gary Conte, on the reorganization.
- Lance Friedmann is senior vice president of marketing for Kraft Foods International Inc.
- Sanjay Khosla joined Kraft Foods in 2006 as executive vice president and president of Kraft Foods International.
Karen May, executive vice president of global human resources, was the executive sponsor, with
Dave Brearton, of Kraft Foods’ reorganization.
Nick Meriggioli is the president of Kraft Foods’ Oscar Mayer business unit.
Rick Searer is executive vice president of Kraft Foods and president of Kraft Foods North America.
Jean Spence is Kraft Foods’ executive vice president of research, development, and quality.
This article was created through a “virtual roundtable”: a series of interviews edited into a single story line. Also contributing to this article was consulting writer Robert Hertzberg. For more information, contact email@example.com.