Case Study Report

This case study was prepared by Jeanne W. Ross of the MIT Sloan Center for Information Systems Research and Cynthia M. Beath of the University of Texas at Austin. This case was written for the purposes of class discussion, rather than to illustrate either effective or ineffective handling of a managerial situation. The authors would like to thank the executives at Campbell Soup Company for their participation in the case study and to acknowledge, in particular, Doreen Wright, Joe Spagnoletti, and Michael Moeller for their tireless efforts to clarify critical success factors and outcomes.

© 2008 MIT Sloan Center for Information Systems Research. All rights reserved to the authors.

Massachusetts Institute of Technology Sloan School of Management

Center for Information Systems Research

Campbell Soup Company: Harmonizing Processes and Empowering Workers

In June 2007, Mark Sarvary, then President of Campbell North America, observed that, over the prior five years, Campbell Soup Company had transformed its systems to enhance business operations. He noted, however, that the com- pany had just started to reap the benefits:

We are most of the way through unifying our computer systems across the whole company. This is a very fundamental change, not only to the systems but to how we operate—how integrated we are and how coordinated we are. The task that we face, starting literally now, is what we’ve called “the benefits reali- zation focus.” —Mark Sarvary Former President of Campbell North America

In the summer of 2007 Campbell Soup was starting the third year of Project Harmony, a four-year initiative introducing common trans- action processes and a more fully integrated systems solution across Campbell’s businesses for transactional activities in supply chain, accounting, and customer services. As the com- pany implemented system and process changes, it was applying lessons learned by the many consumer products companies that had already traveled that path. But management found fewer

proven templates to guide their efforts to fully realize the benefits of a business transformation of this nature:

Unlike the implementation of SAP, where there are binders and books and lots and lots and lots of very direct comparable experience from other com- panies, this [driving business benefits from SAP] is much harder to do. I think it will take two or three years to make the material changes we’re going to need to make... —Mark Sarvary

Campbell management had built the business case for the project based on operating cost reductions, but it was clear that full benefits realization was more than a cost cutting exer- cise. Sustained benefits depended on a more empowered work force working across business and functional lines to improve business perfor- mance. Moreover, the behaviors required in this new environment would have to start with the project effort itself.

Company Background Founded in 1869, Campbell Soup Company was a global manufacturer and marketer of high quality foods and simple meals with 2007 sales of almost $8 billion. Campbell’s 23,000

Ross and Beath Page 2 CISR Working Paper No. 374

employees worked in the company’s Camden, New Jersey, headquarters and 41 manufacturing facilities in 11 countries, as well as sales offices in some of the 120 countries in which Campbell’s products were sold. The company’s operations were divided into four segments: U.S. Soup, Sauces and Beverages (e.g., Campbell’s soups, Swanson broths, V8 juice, Prego and Pace sauces); Baking and Snacking (e.g., Pepperidge Farm cookies, crackers, and bakery items, Arnott’s biscuits and salty snacks); International Soup and Sauces (e.g., international sales of Campbell’s, Prego, Swan- son, and V8 brands, as well as Royco soups and Lesieur sauces in France; Devos Lemmens mayonnaise and cold sauces, and Royco soups in Belgium; Blå Band soups and sauces in Sweden); and Other (primarily Godiva1 choco- lates and Away From Home operations).

When Douglas Conant was named CEO in 2001, he took the reins of a company that was lagging its competitors in both financial and market performance. In addition, Campbell was facing competitive pressures from many sides. Consumers were becoming more price and health conscious. Significant consolidation in the industry meant that Campbell, a medium sized firm, was competing in an industry domi- nated by giants such as Kraft and Nestle. Moreover, Campbell’s upstream agribusiness partners and downstream retail partners were also consolidating and becoming increasingly powerful. And downstream retail partners were frequently competing with Campbell through their private label offerings.

Conant set out to rejuvenate the 132-year-old company. His strategic vision involved revital- izing US soup sales, shifting Campbell’s portfo- lio of brands and products to emphasize growth, and driving a quality agenda. Campbell’s opera- tional strategy, adopted in 2002, was to distin- guish core business activities from non-core business activities, and to then manage non-core activities for low cost while managing core 1 On August 9, 2007, management announced that it was seeking strategic alternatives, including divestiture, for its Godiva business, which, unlike other Campbell products, was sold in department stores and specialty shops.

activities—sales, marketing and R&D (especial- ly retail execution), trade management, and product lifecycle management—for differen- tiation and growth.

By 2007, Campbell was significantly outper- forming industry averages. One key indicator, shareholder returns, found Campbell generating returns of 16.2% compared to the 7.7% earned by companies in the S&P’s packaged food index. [See Exhibit 1 for detailed financial performance.]. This performance was enabled by significant business process and IT improve- ments that Conant and his management team initiated in 2002.

Rethinking IT Doug Conant brought Doreen Wright to Campbell in 2001, as the company’s first corpo- rate CIO. Wright, who reported directly to the CEO, took the helm of an IT organization that had long been decentralized. The resulting systems and infrastructure reflected the local decision-making structure:

There was no glue holding IT together as a global function—none. Every busi- ness around the world had its own IT department and made its own decisions. Little was centralized and there was no global governance. We found ourselves running multiple versions of duplicate systems. This left us in the unenviable situation of being a relatively small company with more than one thousand applications. —Doreen Wright SVP & CIO

An early change established a dotted line reporting relationship between distributed busi- ness unit IT executives and the CIO. That dotted line, which eventually became a solid line, allowed Wright to rebuild the global IT func- tion. Corporate IT leaders focused on gover- nance and architecture, while business unit IT leaders were responsible for relationship man- agement. Wright reallocated the incentives of IT leaders to reflect their role in enterprise-wide, as well as business unit, performance.

In reorganizing and redirecting IT efforts, Wright implemented what Gartner referred to as

Ross and Beath Page 3 CISR Working Paper No. 374

“IS Lite.”2 The IS Lite approach centralized and shared IT services not unique to business units. IS Lite also meant outsourcing IT responsi- bilities that did not differentiate the company from its competitors. By outsourcing the non- distinctive IT services, Wright could focus Campbell IT leaders on strategic requirements, such as relationship management, governance, and architecture.

Wright noted that outsourcing infrastructure operations and other technical responsibilities would not necessarily be a cheaper alternative to running computer operations internally. What the company gained from outsourcing was a partner who could deliver best practice, readily available computing capacity, and protection against disasters. Most importantly, outsourcing added management bandwidth:

I need the capacity of my staff to be focused on what is new, to understand it and to help keep up with evolving needs. With our business people clamoring to do data synchronization and collabo- rative planning with our customers, and introducing new R&D capabilities, and trade promotion capabilities, the last thing I want is to tie up my leaders’ time with is the running of the computers themselves. Everybody’s mind needs to be on how we stay ahead to support the business. So, I completely outsource the infrastructure. —Doreen Wright SVP & CIO

Building a World Class Infrastructure Wright’s starting point for outsourcing the infrastructure was an existing deal with IBM. In 1995, Campbell had outsourced most IT oper- ations and desktop support to IBM. In 2003, Wright worked with key IBM representatives to renegotiate the agreement. The new deal shifted some responsibilities between Campbell and IBM so that Campbell staff handled direct interactions with internal clients (e.g., desktop support) and IBM took on more application maintenance. IBM also arranged a three-in-the- 2 “The Reality of IS Lite,” Gartner EXP Premier, September 2003.

box management team: (1) an account manager who took responsibility for coordinating IBM sales and services and served as the point of escalation for any issues that needed to be resolved; (2) an IBM Business Consulting Services executive who assumed responsibility for Campbell client satisfaction and IBM’s revenue and profit goals, and (3) an executive from IBM’s Strategic Outsourcing business who oversaw planning and execution of the services contract. These three executives coordinated IBM’s services for Campbell, and they often served as Campbell’s general contractor with other vendors as well.

Wright and the IBM team jointly committed to their mutual success, meaning that IBM would provide Campbell with a reliable, cost effective IT infrastructure and world class IT operational processes, while Campbell would recognize IBM’s need for reasonable profitability and revenue growth. These commitments were regu- larly tested. When Campbell was looking for cost savings, Wright turned to IBM to find ways to drive down base costs while still meeting Campbell’s service needs. At the same time, she extended the companies’ relationship by selecting IBM as the company’s integration partner for their SAP implementation. The arrangement benefited both sides, as one IBM executive explained:

We’re driving base costs down for Campbell in strategic outsourcing. But as IBM gets these business value pro- jects, like SAP, they add new require- ments into the system for hosting, and that funnels back to me as growth. This is something that I think Doreen [Wright] recognizes. She came to me in 2004 and said she needed $3.5 million to help fund the SAP project. Well, if I can drive down my cost, and I know Campbell is going to turn around and invest those savings in a way that will benefit our business consulting group, why wouldn’t I do that? —Charlie Carpenter IBM Global Services’ Strategic Outsourcing Executive

Ross and Beath Page 4 CISR Working Paper No. 374

While IBM consistently met its commitment for reductions in annual base charges, the total IBM contract was growing between $.5 and $1 million a year because of new service require- ments. In total, IBM represented a significant portion of Campbell’s IT budget. Wright valued IBM’s contribution to Campbell’s operations, noting that IBM’s staff was at least as qualified and just as reliable as anyone she might be able to hire to manage these activities internally:

You have to be realistic about what you are looking for from your partners. If we were providing our own data center services, we would occasionally make mistakes and bad decisions. It is not different when you have an outsourcer. What is important is that the two sides are each deriving benefits, that they trust each other, and that there’s give and take. IBM has a huge vested interest in this company. They want us to win like we want us to win. —Doreen Wright SVP & CIO

Leveraging a Robust Infrastructure By 2005, Campbell had invested $20 million to make the infrastructure “standard and industrial strength.” Benchmarking indicated that Campbell was average or above average in technical infrastructure, security control and computer operations reliability. By leveraging this robust foundation, IT eliminated $4 million in base operating costs. However, Campbell’s appli- cations software had been developed over the years to address very specific needs in functional and business unit silos. The result was not only a messy applications landscape but also nonstandard business processes:

[In 2005] we had five different companies in North America that sold to our cus- tomers and every one of them did busi- ness with our customers a different way. They must have thought we were dysfunc- tional. We didn’t have one way of check- ing credit, or checking inventory to tell you whether we had it or not, or one way of billing. None of that was standardized. —Rob Austermehle Head of Customer Service Center

With a solid infrastructure in place, Campbell set out to clean up its systems and transform its business processes. The goal was not just proc- ess improvement; management wanted to create a more competitive and agile company:

The big change is that we’re going to have the same software pretty much in every unit. And we’re going to be fanat- ical about reporting everything the same way, handling transactions the same way. Changes to software and process are going to be controlled centrally as opposed to de-centrally… I think it’s going to be much more efficient. People are going to spend more time thinking about how to utilize information to their advantage rather than changing the way it’s rolled up or what line it appears on. —Robert Schiffner SVP & CFO

Mapping a Transformation At a senior management meeting in 2003, Steve Smith, the IBM Business Consulting Services executive at Campbell, presented an “art of the possible” business case that highlighted the value of more standardized business processes across Campbell’s businesses. He argued that an effective SAP implementation would support more standardized and integrated businesses and help position Campbell for future growth.

CFO Bob Schiffner endorsed the potential cost savings and CIO Doreen Wright estimated that implementing SAP would eliminate hundreds of complex applications. CEO Doug Conant quickly became an enthusiastic proponent of a business transformation built around a global implemen- tation of SAP. The transformation initiative was dubbed Project Harmony.

Conant and other senior management team members did not believe that Campbell would be able to implement its strategic vision without changing the way it did business. Management allowed four months for development of a formal business case and implementation plan before making a proposal to the board.

Ross and Beath Page 5 CISR Working Paper No. 374

In May 2004, Campbell’s board committed $125 million in capital for a three-year project that included implementing SAP in North America.

Planning for Change Before embarking on Project Harmony, Campbell engaged its senior managers to define the parameters for the project and discuss how the company should operate going forward:

We pulled people in from around the world and we asked, “What things have to be the same when we deploy SAP?” It was a process to determine what had to be common and what could be unique by business or region. It was also a way to engage everyone in the fact that we were going to deploy SAP around the world over time. —Doreen Wright SVP & CIO

Senior executives developed a set of operating objectives which Campbell referred to as the Global Framework Objectives:

Utilize standard SAP capabilities (minimize development and maintenance cost)

Reduce unnecessary touches (management by exception)

Maintain/Improve customer service (“the perfect order”)

Maintain/Enhance order fulfillment process productivity (KPIs)

Utilize “available to promise” across all business units

Create customer transparency Drive “easy to do business with” concept

(flexibility) Streamline/Improve controls Present a single voice to customer

The published Global Framework objectives helped to focus the efforts of the Project Harmony team:

So those were the high level guidelines, and we pasted them on the wall when we started Harmony to say, whatever we do over the next two or three years, we don’t want to lose sight of these visions and these guidelines so that we don’t put

a roadblock to something that we might want to do five years out. —Rob Austermehle Head of Customer Service Center

Project Harmony focused on three work streams: Make-to-Ship, Account-to-Report, and Order-to-Cash. These work streams flowed horizontally across the company’s businesses. Standardizing and integrating these work streams offered the potential for process effi- ciencies and improved customer service.

The extended debates leading up to the Global Framework helped to clarify the opportunities of a common process solution and to solidify long- term commitment at the senior management level. Throughout the life of the project, senior managers reinforced the goals defined by the Framework:

When my global team gets together, we have an SAP review each time, and Nigel [Nigel Payne, project lead for the make to ship work stream] updates us on the issues that are being worked by the teams… The discussions at the global team level are generally more principle- based. We make sure the direction we are heading is consistent with our goals by reinforcing decisions like, “We are going to have one standard solution, and we’re not going to deviate. And even though it’s not quite what you want, the standardization benefits are going to outweigh the loss.” —David White SVP GSC

Because managers at Campbell had traditionally focused on business unit performance, the prin- ciples captured by the Global Framework high- lighted the need for a new mindset emphasizing enterprise-wide performance. Management adopt- ed Total Delivered Cost (or TDC, as they referred to it) as a key performance metric to focus management attention and gauge progress on enterprise-wide processes. TDC encom- passed the end-to-end cost of producing a product and getting it to the customer. Campbell’s goal was flat TDC, and flat TDC required thinking well beyond the boundaries of a single function or business:

Ross and Beath Page 6 CISR Working Paper No. 374

Total delivered cost is the total cost of making a product and getting it to the consumer. We target TDC to be flat, which means that the full cost of producing and shipping something to the consumer will be the same this year as it was last year. We essentially have to improve our productivity by the same amount as inflation goes up. SAP gives us both the ability to know TDC and the ability to improve it. —Mark Sarvary Former President of Campbell North America

While SAP was the biggest part of Project Harmony, the full scope of the project included additional systems components (e.g., creation of a data warehouse) and organizational design components (e.g., the establishment of a shared services organization for finance and account- ing). Due to the magnitude of the change the company was undertaking, Campbell invested in a six-month planning stage to identify key resources and prepare managers at all levels for the transformation:

During the project planning phase, we really defined the business case, the implementation plan, resource require- ments, etc. And that took quite some time, actually. We invested a lot of time, and that’s what I feel made the differ- ence. We drew up charts showing our messy legacy as-is information systems [Exhibits 2a and 2b], our integration points, etc. We clearly delineated the scope for the Harmony program, the process and function scopes, the objec- tives of the program and ultimately the business case. —Roberto Depani VP, Project Harmony

Organizing for Change During the planning stage, Campbell designed a three-pronged governance structure to ensure smooth delivery and rapid realization of project benefits. Key decision making bodies were (1) a sponsor team comprising senior executives, (2) an operating committee made up of project leaders, and (3) three process advisory groups— one for each of the three key processes.

Campbell’s governance structure was intention- ally heavy with senior leaders to ensure success- ful implementation. [See Exhibit 3 for a descrip- tion of the project structure.]

Senior Management Sponsors Initially, three senior executives sponsored Project Harmony: CIO Doreen Wright, CFO Bob Schiffner, and President of Campbell North America Mark Sarvary. These three executives, along with Steve Smith, the IBM partner client executive on Project Harmony, established project expectations and goals. Later, David White, Senior Vice President and Supply Chain Officer, became a fourth member of the sponsor team.

At their bi-weekly meetings the sponsors re- viewed progress and provided resources to ensure Project Harmony met targets. All requests for deviations from standard had to pass through this team—a requirement that severely limited the number of exception requests. The sponsors also considered projects and other activities that had to be taken off the table in order to maintain focus on Project Harmony implementation.

Operating Committee The operating committee ran the project on a daily basis. Roberto Depani, an IT leader with global experience, was named Vice President in charge of Project Harmony. He led the operating committee, which also included Michael Moeller, Vice President, Corporate Program Office, who was responsible for change management, a technical lead, the IBM project director, and three senior managers, each accountable for one of the three global processes.

All three process leaders had extensive opera- tions experience. Lon Alness, head of Account- to-Report was a former VP of supply chain finance. Rob Austermehle, who had been the VP in charge of Campbell’s customer services center was assigned to lead Order-to-Cash. And Nigel Payne, the Make-to-Ship process lead, was former VP of procurement and had run a manufacturing plant.

Ross and Beath Page 7 CISR Working Paper No. 374

The operating committee met weekly to make decisions on the interdependencies among the process areas as well as ensuring the overall program remained on track.

Process Advisory Groups Three process advisory groups (PAGs) advised the process teams. The PAGs were chaired by senior executives. For example, David White, Senior Vice President of Global Supply Chain, headed the make-to-ship PAG; Denise Morrison, then Chief Customer Officer, chaired the Order-to-Cash PAG. Although most of the basic process decisions were left to the process teams, the process advisory groups helped with design when the project leaders needed input, and they reviewed proposed end-to-end process- es to identify issues. As Rob Austermehle explained, “The 20% [the global process team] couldn’t decide went to the PAG.”

The Project Team The project team consisted of approximately 60 Campbell people and more than 70 consultants and other external experts. The team moved into a separate building and focused full-time on project implementation:

The Campbell people were either back- filled or their areas were restructured. We formally took them out of their cost center and put them into my cost center. So, they now report to me and their costs flow into the SAP project. This allowed the team to stay focused exclusively on the transformation effort and not on their pre-existing jobs. Having a core team of dedicated project leaders was a critical success factor. —Roberto Depani VP, Project Harmony

The three process leaders recruited top subject matter experts for the global process teams. As a result, the global process teams had both lead- ership experience and process expertise:

We built a team around people from the business who really understood it. So, the three of us [the process leads] could get in a room and come up with a pretty good proposal for Roberto or for the

business, to say, “This is how we think it should happen.” We’d bounce that against our process owners on our indi- vidual teams and when we couldn’t de- cide, we got the right people from our three teams in a room and locked them down until they came up with a solution. —Rob Austermehle Head of Customer Service Center

Specialists from IBM and SAP supplemented the efforts of Campbell’s employees. IBM had more than 70 people involved in Project Harmony and represented approximately 20% of the total project budget. Fifty members of the IBM staff were doing offshore development, mostly in India. Led by IBM’s Project Director, John Terzis, IBM staff provided project management expertise and SAP configuration expertise, in addition to supporting development efforts:

I have accountability for the technology, for configuration, for the programming, which, by the way, is largely IBM but not all IBM. —John Terzis IBM Director for Project Harmony

SAP provided primarily consulting services. Both the SAP engagement manager and the SAP account manager worked directly with Doreen Wright and Roberto Depani on a regular basis to develop and review strategy and the implementation roadmap. The SAP engagement manager frequently sat in on operating commit- tee meetings to alert management to potential issues. SAP also provided a lead consultant to each of the three work streams. Campbell man- agers considered IBM and SAP to be strategic partners and insisted that the two vendors work together as strategic partners as well:

The IBM—SAP—Campbell’s partner- ship has been successful as a result of transparency, clear delineation of roles and responsibilities and the assignment of “A” players in all key positions. —Doreen Wright SVP & CIO

Ross and Beath Page 8 CISR Working Paper No. 374

Managing for Change Michael Moeller, Campbell’s Vice President heading change management efforts, initiated communications and education about Project Harmony almost immediately after the project was announced. His role involved winning necessary commitment from project team mem- bers and key stakeholders to the goals of the program and ensuring that Campbell had in place the needed skills, structures, behaviors and attitudes to enable a new way of doing business.

Early on, Moeller and project leadership con- ducted several all-day meetings with the project team to make sure all members understood both what they were fundamentally trying to ac- complish and to explain the methodology and governance they would rely on to succeed. One of the objectives of the meetings was to create a single team identity that avoided unnecessary distinctions between consultants and Campbell employees or among the different work streams.

Change management efforts also focused on senior leaders to make sure they understood the benefits and implications of greater process integration and standardization. One of those leaders, Michael Dunn, described the senior management communications:

[About 18 months before we went live] at a Global Leadership Team leadership conference that Doug Conant conducted, we learned that SAP was going to be one of the key strategies to help drive excel- lence within Campbell’s. So we knew about the project well ahead of time. With my colleagues from the other thermal plants, we started discussing the project and what it was going to take to execute this new system. We were trying to get ourselves up to speed, educate ourselves, and to get our minds around what it was going to take to be successful, to actually implement. —Michael Dunn Plant Manager, Paris, Texas

Moeller noted that it was difficult for an organ- ization that had traditionally operated in silos to absorb the full implications of the changes prior to deployment. So he took a “peeling the onion” approach, starting with broad goals and high

level expectations and drilling down to work group and individual level implications as the path became clearer. Much effort went into mak- ing sure site managers understood the new roles and responsibilities emerging in the company:

We created functional roles, which described the different sets of responsi- bilities that somebody would need to perform in each process. These roles were used to determine individuals’ se- curity access in the system and what training classes they would take, be- cause we did role-based training. So role assignment was a really critical process for us. Managers and employees needed to understand the key process changes and what it meant to individual responsibilities. —Michael Moeller VP, Corporate Program Office

Project subject matter experts first conducted management workshops to discuss key process changes and explore their implications.

Subsequent organizational alignment sessions examined how the sites would define individual roles:

The project subject matter experts would explain each role while emphasizing what was new or different from how we were then structured in the business. Then we’d turn to the relevant business managers and say, “OK, who makes sense to perform that set of activities given our new requirement?” We liter- ally had a list of names from the relevant business groups on the wall, and after some discussion we’d go through the list and put an X in the box for everybody for whom the role made sense. After we assigned all the roles, we’d go back and look across each individual’s set of assignments to make sure we hadn’t over- subscribed anybody with too many roles, especially because people could still have important responsibilities outside their process roles. —Michael Moeller

Site managers then took responsibility for gain- ing buy-in to process and role changes and for filling any skill gaps that wouldn’t be addressed

Ross and Beath Page 9 CISR Working Paper No. 374

by the formal training program. Project team leaders worked with site mangers to overcome natural resistance to change and create excite- ment about new opportunities:

There is a layer of management in parts of this organization who have built ca- reers and survived because they are the glue in the legacy application landscape. They know how to get information from one system and provide it either to other systems or as reports. So, they were actually our biggest challenge and were occasionally resistant to change because they recognized that their jobs were going to change significantly. A big part of my role has been persuading those individuals that they can now begin to add strategic value through analysis of the data rather than creation and manipulation. —Nigel Payne VP Global Supply Chain Process Excellence

In working with employees at all levels of the organization, project leaders emphasized that the changes would be transformational:

We’ve gone into the plants and said, “As from go-live, it is as if you have joined a new company, you have to forget every- thing you’ve been doing for the last however many years.” —Nigel Payne

Moeller emphasized to team members that they should focus first, on getting new systems and processes adopted successfully, and second, on quickly generating benefits. Despite all the advance efforts, many of the changes needed for benefits realization would take place after implementation:

I used to say to my team that our goal right now isn’t to get on the medal plat- form. We need to focus on getting suc- cessfully across the finish line first. Then, once we have some direct experience with our new system, we can use post implementation activities to refine and optimize our approach in order to capture the full benefits of Project Harmony. —Michael Moeller VP, Corporate Program Office

Deploying SAP In April 2006, Campbell’s Canada’s head- quarters and two Canadian plants served as a pilot for the broader SAP implementation. Canada was a microcosm of the much larger US business and would implement about 75% of the project’s global processes.

As much as a year before implementation, Campbell’s Canada business created a Deploy- ment Committee, comprising a subset of the Canada business’ senior leaders. The Deploy- ment Committee reviewed Project Harmony requirements in the context of the master planning process. Recognizing the magnitude of the changes and the heavy resource require- ments, the Deployment Committee staged or killed most other change initiatives, including product rollouts, pricing changes, and new promotion efforts:

Our businesses recognized early on that tough choices would have to be made. It was clear that we had to focus local resources on Project Harmony so Canada’s Deployment Team didn’t try to do much else, other than keeping the business running. —Michael Moeller

In addition to the employee preparation, the Canada team worked with the global project team to prepare customers for the rollout:

We were not going to fall into the trap of disrupting our external customers as a result of this. We got out in front a year early, talking to our customers about it, laying out a game plan, using regular meetings that we had with them to update them on the status of the project. And we have written evidence from cus- tomers that they felt this was a best in class implementation. In fact, we had some customers call after implemen- tation and say, “did you go live?” and that was a real tribute to the team. —Rob Austermehle Head of Customer Service Center

Site preparation involved a several month period of team building, data cleansing, local config- uration tailoring, application interface develop- ment, and organization design. Each site

Ross and Beath Page 10 CISR Working Paper No. 374

received three to six weeks of training before go-live.

During the conversion, twenty or more local super users supported 450 hands-on users. The project team provided intensive support:

After go live, we enter at least a month to two months of what we call “hyper care.” It’s all hands on deck. Two people doing one job. The whole core team, the support team, etc., are all focused on the particular site, providing hands-on support, resolving things as they come up. —Roberto Depani VP, Project Harmony

Subsequent rollouts benefited from the learning acquired in prior implementations. Before Campbell’s world headquarters and Paris, Texas plant went live in February, 2007, senior man- agement and team leaders visited the Canadian site to learn from their experiences.

We walked through all of the key learn- ings from their deployment focusing on staffing, leadership and change man- agement. We spent an entire day going through the good, bad, and the ugly. It was a transformational moment in the program that brought to life for our leadership team just how involved they needed to be and the focus it would de- mand versus taking on other day to day business programs. —Joe Spagnoletti VP, Information Technology

Senior leadership emphasized that the goal of every manager was not just the success of the current implementation but also the success of the next one:

Part of your job is not just to roll out SAP in your business but to help the next plant or the next business get ready. So, Canada’s job didn’t end with their go live. Their job is then to help Paris fig- ure it out. And Maxton [North Carolina plant] has gone to Paris to learn from them. —David White SVP GSC

Recognizing that they could learn from prior implementations, business leaders sought out their experienced colleagues. Prior to the Pepper- idge Farm headquarters rollout in November 2007, management asked Joe Spagnoletti, who had led the headquarters rollout, to do an assessment of their readiness for rollout:

I met with senior leadership, interview- ing them one at a time, to assess their awareness and readiness to drive change in the organization post go-live. This assessment revealed that the Pepperidge Farm team had done an exceptional job in many ways embracing that which Campbell USA had learned, and building upon that experience when developing their plans. As a result of this linkage and open dialog, their communi- cation and change management plans further improved upon those which were used previously. —Joe Spagnoletti

Senior management also continued to empha- size the goals of Project Harmony. As each new site prepared for rollout, local managers would find parts of the SAP template that were not consistent with the way they did business. Project leaders were willing to change the template only if the standard process was not consistent with the way they could do business:

It was not unusual to be approached and be asked, “Why are we doing it this way?” or “Why wasn’t I included in the design of the new process?” and the answer was, typically, “It’s been decided by the core team process leads and we’re going to do it this way at the global level. Now let’s talk about how we’ll adopt that practice.” —Joe Spagnoletti

By the end of 2007, Canada; Campbell’s USA plants in Paris, Texas, and Maxton, North Carolina; Pepperidge Farm; and Campbell’s headquarters had gone live. [Exhibit 4 shows the project timeline.] The rest of Campbell’s US plants would be live by the end of 2008. As each new site learned from earlier sites, project leaders were starting to move into new roles that would focus on benefits realization.

Ross and Beath Page 11 CISR Working Paper No. 374

Benefits Realization Early business cases for Project Harmony had cited potential benefits, largely in supply chain, finance and accounting shared services, and back office rationalization. Management knew that benefits would be limited until North America had fully implemented, but they expected that return on investment would be around 20%.

We built an overall financial model of benefits flow that distinguishes between direct and indirect benefits. Certainly there are some clear-cut direct benefits that can be directly tied to what we’re doing. These can be measured, tracked, monitored and managed for delivery. There are also other benefits that you know the project is bringing to the busi- ness but are harder to get your arms around, because they are more aspira- tional or an indirect result of doing the project. —Roberto Depani VP, Project Harmony

By late summer of 2007, Campbell reported hard savings in IT from retiring legacy technol- ogy and applications and reducing maintenance:

We identified substantial ‘hard’ savings for IT and we achieved them. We’ll be just under $10 million this year [2007], and next year we’ll be over $10 million. —Doreen Wright SVP & CIO

The company had also started generating savings from centralized accounting and finan- cial services. Rob Austermehle, whose work as head of the order-to-cash work stream was nearly completed, became head of finance and accounting shared services:

[Shared services yields] hard benefits. I don’t need to have duplicate operations for accounts payable in five different businesses when we actually share many of the same vendors. —Rob Austermehle Head of Customer Service Center

An early source of savings was coming from effective use of transparent information where SAP had been implemented. Campbell’s had significantly reduced ingredient losses as a

result of process discipline and better visibility to critical information. Key to such improvement was empowering workers to use newly available real-time information in their daily decisions.

Organizing for Benefits Realization To drive supply chain benefits, David White, the Senior Vice President of global supply chain, had assigned process owners to create structure around major processes, such as global reliability and produce-to-demand. For example, Nigel Payne had moved from his role as the make-to-ship work stream leader to head of a supply chain center of excellence (COE):

Everyone realizes there is now only one way of doing something. They now know that if they want to change anything, if they have a better way of doing some- thing, it has to be analyzed, it has to be assessed against all the other businesses, before it can go forward. So that’s part of the COE’s role. Another is building up knowledge networks within the organization. My future role is to encourage and spread the knowledge network, but also provide the gover- nance over the processes. So, if anybody wants to change anything within the system, it has to come back to the process governance role. —Nigel Payne VP, Global Supply Chain Process Excellence

As plant managers identified opportunities to improve processes, they naturally shared their learning across the organization. Nigel Payne’s center of excellence would accelerate that process:

If you go up to Canada, the plant manager can tell you 20 places where newly available information can save money, but it wasn’t obvious four months ago. They’re figuring it out as they go. Well, now Canada is training the Paris and Maxton plant managers, so they don’t have to spend four months figuring it out... Nigel is going to develop a process that allows us, as the learnings come out, to spread them quickly across the 22 plants—David White SVP GSC

Ross and Beath Page 12 CISR Working Paper No. 374

Absorbing a New Culture In Campbell’s 2007 annual report, Doug Conant reported that the Campbell Soup Company was “well on our way to realizing our mission of building the world’s most extraordinary food company.” By the end of 2008, all of North America would be running SAP, the project team would be fully absorbed back into the business, and the last IBM consultant would have rolled off the project.3 When North America was completed, Campbell would start to install its SAP footprint overseas:

We will experience new challenges as we go overseas. For our North American implementation, many of the Harmony team members were local and were people we had worked with in the past. There was a significant degree of famili- arity and trust which made it very easy to accept change and work through dif- ficult decisions. As we implement the program in other regions, there will likely be cultural differences in ap- proaches and styles. We will have to work to develop relationships and build the trust necessary to be successful. In addition there will be significant phys- ical distance and time zone differences between the core project team and the business implementation teams. We will need to adopt new approaches for staffing and communicating for this phase of the program. We won’t be able to walk the hall and tap people on the shoulder. —Joe Spagnoletti VP, Information Technology

Despite the challenges they faced as they deployed regionally and ultimately globally, Campbell management had decided to accept the benefits and constraints of relying on SAP for its core systems:

We still haven’t quite got the cultural shift to realize that we are an SAP shop and my challenge going forward is, “Why shouldn’t the proposed new

3 However, as part of their infrastructure services contract, IBM retained primary responsibility for application management services related to SAP.

solution be SAP? Why do you want a best in breed that may cost a lot to interface?” The biggest challenge we’ve had for the last two years is the inter- faces between SAP and retained legacy applications. And I don’t think people really realize the cost, the true cost of a lot of these interfaces in terms of on- going costs, maintenance, error mes- sages, fixing, re-fixing. SAP has got lots of areas to develop still, but there’s no reason why in five years time we couldn’t be running this business better with literally nothing but SAP. —Nigel Payne VP Global Supply Chain Process Excellence

ADDENDUM (MAY 2008) In June 2008, Campbell was on track to complete the North American implementation of Project Harmony by the end of the year, as written into Project Harmony plans almost four years earlier.

The governance process at each deployment site was designed to drive the focus and decision making required for successful implementation without disrupting the business. It enabled busi- ness leaders to make the necessary trade-offs to sustain business performance while putting in the new system and processes:

It is significant that we did not relax performance expectations during imple- mentation. The businesses didn’t get a pass on their numbers just because we were putting in SAP. Instead, they stayed focused and made the choices they needed to make to run the business and support the project. Remarkably, each of our businesses was able to exceed performance expectations during the period we were implementing the system at its headquarters location. —Michael Moeller VP, Corporate Program Office

Employees at Campbell’s North American facilities noted that SAP provided valuable information to support their efforts. For exam- ple, customer care employees could manage by

Ross and Beath Page 13 CISR Working Paper No. 374

exception; they followed up on fewer order problems while responding more effectively to orders needing their attention. Supply chain employees reported the ability to quickly diag- nose errors and provide training when errors resulted from people mistakes. Plant managers claimed significant savings from faster recogni- tion of equipment problems. And one controller found that process improvements had reduced the control points for Sarbanes Oxley from more than 60 to around 20. Campbell management was encouraging individuals throughout the firm to accept ownership for generating benefits from SAP:

Individual employees can begin to build their own foundation and begin to un- derstand that their job going forward is to evolve the solution, ask questions, and come up with ideas on how to better use this toolset. —Charlie White Director of IT, Project Harmony

The firm was also realizing benefits from ongoing employee collaboration. An inventory supervisor in Maxton, who learned how to use additional reports to check warehouse inventory levels, quickly shared her learning with the Paris, Texas, facility, and Paris passed on the learning to Canada. Employees throughout the

company were identifying ways to reduce total delivered cost (TDC):

It never dawned on us that what we were doing was empowering thousands of workers, but that is the effect of pro- viding transparent information and the authority to use it. —Doreen Wright SVP & CIO

Individuals reported thinking about how their actions affected downstream processes. They described a more “integrated” business that called for looking outside their local facility. Employee willingness to think about how their actions affected the entire company instead of simply “doing my job” had fundamentally changed Campbell:

The most profound thing that has hap- pened here is not the systems change or the business process change. The most profound change—and it was profound— was the culture. —Doreen Wright

In May 2008, Campbell management was pre- paring for its next deployment in Australia. In doing so, leaders were assessing the extent to which the company’s cultural change could be replicated in other parts of the world.

Ross and Beath Page 14 CISR Working Paper No. 374

Exhibit 1 Campbell Soup Company Selective Financial Data

Beath and Ross Page 15 CISR Working Paper No. 374

Exhibit 2 (a) Legacy Systems Map at Campbell

Ross and Beath Page 16 CISR Working Paper No. 374

Exhibit 2 (b) “To-Be” Systems Map at Campbell

SAP

Mini MIINT

Supply Chain Repository

Manugistics

Network Strategy

Accusort

Telerx

PLM

Prosync

Xerox

DataSync

Trendset CASS

3rd Party

Depot HHC Gateway

SDA HHC Gateway

EDI Translator

First Union

BMO

Harris Bank

Wachovia

CIBC

Excel Spreadsheets

Carolina Reclaims

Customer Workbench

GL Error Correction

Canadian T&E Unclaimed Assets

Asset Recovery

Peoplesoft Khalix Hyperion

PF SDAs

Invoice PDFs

PUC

Financial Workbench

IQ2

GTM

Wheels

Gelco

CMS Coupons

i2

CWMS

Brokers

VIP

PavingCRP

FreitRater Customs

Compliance Export Receipt

Exception

FSN Kovis EnvisioniSales

Anacomp

RMS

Wonderware

eMAA Paybase

Futurecast

Plant Floor Applications

TIPS

Kronos

CH Robinson

Customers

PF IRISPF Thrift

Beath and Ross Page 17 CISR Working Paper No. 374

Exhibit 3 Project Harmony Governance Structure

Project Harmony Governance Model*

PAG** PAG**PAG**

CUSA Deployment Committee

Pepperidge Farm Deployment Committee

Order to Cash Make to Ship Account To

Report

Technology Team

Change Management Team

Program Management Office

Sponsor Group

Operating Committee

Deployment Teams

* Project Harmony governance model simplified for publication. ** Process Advisory Group (“PAG”) for each functional area.

Canadian Deployment Committee

 

SAP Program Decision Making Model

Project mgmt

Basic design & process/ control

changes w/in process area

or workstream

Major process/ control

changes w/in process area

Cross- process/ cross

workstream

Deployment decisions

Business unit organization & policy changes

Company organization & policy changes

Scope/ budget change

Sponsor group Notify Notify Decide

1 Notify/ Decide2 Notify Decide

3 Decide4

Process advisory group

Notify Notify Decide Consult Notify Notify Notify Consult (scope)

Operating committee Consult Notify Notify Consult Notify Consult Consult

Program management office (PMO)

Identify/ Recommend/ Decide

Notify Consult Identify Recommend Recommend Recommend

Project team (process area teams)

Notify Identify/ Recommend/ Decide

Identify/ Recommend

Identify/ Recommend/ Decide1

Identify Recommend Identify Identify

Business Deployment Team

Notify Notify Notify Notify Decide2 Decide Consult Notify

1 Basic cross process decisions will be resolved through cross process teams as assigned by PMO; major cross-process decisions with alternate viewpoints will be resolved by the sponsor group.

2 Sponsor group decides issues impacting overall deployment approach and schedule; each deployment site makes master planning decisions related to business activities and deployment/cutover activities within the overall schedule and project requirements.

3 May require approval of CEO for major organization or policy changes. 4 Major scope or budget changes may require board of directors approval.  

SAP Program Decision Model Definitions Decision Roles

Identify – identifies need for decision to be made Consult – provides input to help determine preferred course of action Recommend – gathers facts, consults where appropriate, evaluates alternatives and determines preferred course of action Decide – evaluates alternatives and makes final determination of approach taken Notify – is informed of decisions that have been made

Decision Types Project management – decisions relating to the running of the overall program, including: staffing, internal budget allocations, project standards, policies, protocols, tools, etc. Basic design and process/control changes within workstream – process design and configuration decisions that would not significantly affect the way the business is run or alter how transactions are processed within a particular process area/ minor process control changes Major process/control changes within workstream – process design and configuration decisions that would significantly affect the way the business is run or alter how transactions are processed within a particular process area/ significant changes to process controls Major cross-process/cross-workstream decisions – process design and configuration decisions that have significant impact across multiple process areas Deployment – decisions relating to the implementation of the system within a particular business unit including: deployment staffing, launch sequencing and dates, go/no go, etc. Business organization/policy changes – structural changes within a given business to adapt to new process flows and resource requirements post-implementation/BU level policy changes Company organization/policy changes – structural changes at corporate center or across multiple businesses to adapt to new process flows and resource requirements post-implementation/company- level policy changes Scope/budget changes – decisions to expand/ contract scope of program and any changes to budget that result

Ross and Beath Page 18 CISR Working Paper No. 374

Exhibit 4 Project Harmony Time Line

Remaining Roll Out Schedule

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2

FY05 FY06 FY07 FY08

North America Blueprint

Project Prep.

Canada Implementation

US Blueprint Update*

WHQ, Paris Implementation

May 1

PF HQ, Downingtown implementation

Nov 1

March 1

Q3 Q4 Q1 Q2

FY09

Staggered Plant Implementations

Note: Fiscal year is August 1st to July 31st.