Like most leading makers of mobile phone handsets, Motorola, Inc. understands customer segmentation very well. Managers in Motorola's Mobile Devices business know that the technology-savvy consumer will not buy the same handset as the fashionista or the everyday communicator, so the company offers over 120 models globally.

The difficulty comes in knowing how many of which models to make and sell. Accurate replenishment to retailers' shelves is critical. If a retailer is out of stock of the blue candy bar-style handset that a consumer has her heart set on, she will more likely purchase a competitor's product rather than Motorola's black flip-top phone. In such circumstances, there is a real risk that Motorola loses that consumer for life, not just for the length of the next service contract.

Accurate forecasting is further complicated because one phone model can have multiple SKUs (GSM, color, dual mode, etc.). On top of that, rapid new product introduction adds complexity: Cell phones have life cycles averaging little more than a year, with some as low as six months.

Industry statistics indicate that 50 percent of stock-outs in the consumer goods industry are due to poor retail ordering and replenishment/forecasting processes, 25 percent are caused by overall demand planning and upstream shortages, and 25 percent by inadequacies of in-store processes such as shelving and replenishment.¹ While these statistics are similar for the mobile handset industry, the out-of-stock rate for mobile handsets can be many times larger than for typical consumer packaged goods. Accurate forecasting is critical in this industry because the cost of stock-outs can have a significant negative business impact. Approximately one half of all stock-outs result in lost sales.²

Prior to its move to collaborative planning, forecasting and replenishment (CPFR), the Mobile Device division's sales were highly variable and were not synchronized with customer demand. Motorola had visibility only to shipments into retailers' distribution centers, but not shipments from the retailers' distribution centers to the stores. Forecast error was very high, resulting in excessive stock-outs. Motorola recognized that it needed to improve its collaboration with customers and forecast ability in order to compete effectively in its highly price-sensitive market.

In 2001, the handset division turned to CPFR to improve sell-through performance with its retailers. To head up these efforts, the top management team brought in Jerry Cederlund, who had experience with CPFR in Motorola's messaging business as a customer operations director in North America. (Jerry Cederlund is a co-author of this article.) After demonstrating the potential of CPFR in a manual collaborative pilot with one of the division's retail phone customers, Motorola launched a major CPFR program in mid-2002, accompanied by an organization-wide shift to customer-focused operations teams. These teams were organized around major customers such as cellular operators and consumer electronics retailers.

The goals were to halve forecast error, achieve a 30 percent drop in its inventory, cut channel inventory by 50 percent, and improve on-time delivery by 30 percent. To meet its goals, the division's managers knew they had to share with their retailers their real-time data and plans including forecasts, inventory, sell-through (that is, sales to retailers' shelves, not to end consumers), promotions, product plans, and exceptions.

CPFR is a business practice that improves accuracy by combining the intelligence of multiple trading participants in the planning and fulfillment of customer demand.³ The practice got its start in 1995 as a pilot program between Wal-Mart and Warner Lambert called CFAR, for “Collaborative Forecasting and Replenishment.” In 1997, the Voluntary Interindustry Commerce Solutions Association (VICS) developed the CPFR model and in 1998 published the first CPFR guidelines. Since then, according to the VICS website on CPFR3, more than 300 companies have implemented the process. As these companies worked with CPFR, shortcomings were identified and innovations were developed that were incorporated in a 2004 major model revision by the VICS CPFR committee.

Over time, the VICS initiative reported benefits of CPFR that included in-stock percentage improvements of two to eight percent for products in stores, along with inventory reductions of 10 to 40 percent across supply chains. However, several barriers to adoption have come to light, such as difficulties with real-time coordination of information exchange, substantial investments of setup time and personnel, CPFR's process-intensive nature, its lack of scalability, and the need to synchronize changes between participants. Motorola also recognized that moving to CPFR requires fundamental changes in the way in which it interacts with its participants. It felt that with a successful implementation of CPFR, it could link category management to supply chain planning and execution processes, resulting in increased product availability while simultaneously reducing transportation, logistics and inventory costs. Motorola recognized that the successful achievement of these benefits rested on the coordinated changes to the organizational structure and business processes of both participants.

A common weakness of CPFR efforts is an over-emphasis on replenishment. While streamlined replenishment is indeed the goal, effort toward planning and forecasting is time well spent. Many CPFR projects fail due to lack of executive support, lack of collaboration rigor or because of unclear objectives at the outset. Pilot projects often stall because they cannot scale to work consistently with many of the company's key trading participants. That's especially so when they involve many items and locations, high degrees of required automated processing and lots of exception management. An analysis of this multi-year project reveals significant learning among the managers at Motorola. (See Exhibit 1.)

Motorola's CPFR initiative turned out to be a time-consuming, painstaking endeavor. CPFR requires fundamental changes not only in the nature of the relationship with the retailers, but also in the way each participant conducts business. This collaboration required Motorola to develop new business processes, redesign its organization and adopt systems to support real-time information.

Most CPFR case studies (e.g. “Sears-Michelon” and “West Marine”) describe initiatives that are driven by retailers. However, in this case it was Motorola that initiated the adoption of CPFR with its downstream retail customers. Past experiences of supplier-retailer information linkages have reported varying benefits when suppliers have made the first moves. For instance, a previous study has reported that benefits depend on suppliers' levels of investment in business processes and in domain knowledge that is specific to a supplier-customer relationship.

Supplier-driven initiatives often require the suppliers to make substantial investments to set up linkages such as Electronic Data Interchange (EDI). In Motorola's case, an EDI infrastructure already was in place. Such initiatives also favor suppliers that are motivated to initiate major internal cultural change. For example, it takes true collaboration to push back on a customer that is ordering too much. When this first happened at the handset division, the customer (good-naturedly) queried whether the manager who suggested these order reductions really worked for Motorola!

Traditionally, suppliers and buyers in most supply chains prepare independent demand forecasts. Suppliers have a limited view of the buyer's markets while buyers lack a broad view of category or market insights as well as of production and capacity planning. This often leads to excess inventory and sometimes adversarial relationships. For example, in the semiconductor equipment supply chain, suppliers often penalize buyers for inflated forecasts by delaying order delivery, while buyers may penalize suppliers with a history of poor service by providing them with overly inflated forecasts.

It is vital that the participants are strategically aligned from the get-go. An interesting reflection of Motorola's commitment to that concept: A core business operations team and its customer-focused operations teams (one of which was headed up by Jerry Cederlund) employed both Motorola and customer metrics such as inventory and sell-through in its goals and performance plans. And all the customer team members had the division's retail customers' names on their shirts!

It is also crucial to have strategic alignment within Motorola and the customer organization—alignment of process, organizational and technology strategies with collaborative business strategies. This implies that the business that is advocating CPFR implementation—in this case Motorola—must refine its business strategy to focus on collaboration, first deciding what information will be shared with its customers and then sharing the strategy and getting the customer's agreement to it.

There must also be an internal alignment of the process, organizational and technology factors, all of which support and are in sync with the core business strategy. Specifically, organizational and process changes must be made concurrently with the acquisition and implementation of technology tools that will support more effective information-sharing. All three factors must also scale since the early CPFR activities will eventually grow in scope and range. And for its part, the customer must also see to changes in process, organization and technology factors within its four walls, with full support from its own executive team.

Realignment of the business strategy and the three supporting factors merit a closer look:

Fundamentally, CPFR requires a change in business strategy from the typical “arm's length” transactional relationship with a customer to one that emphasizes collaboration. Common goals must be established and the supplier and customer must reach agreement on what information they will share. It was not an easy discussion for Cederlund and his team to lead. Motorola's handset division had always produced phones that customers really wanted, and its attitude to collaboration reflected that fact. Changing the culture would take many months for each customer.

Using the “Roadmap to CPFR Implementation” that it had developed early on, Cederlund's team began working with the selected North American retailer to agree upon appropriate service levels, metrics and a plan for continuous improvement. To properly qualify the retailer, the team considered the retailer's culture as well as its relationship with Motorola. This particular retailer's culture was not antagonistic toward suppliers but focused on the processes and the causes of market share loss resulting from poor delivery. Motorola had to convince the retailer that a long-term collaborative relationship would be in its best interests by demonstrating short-term improvements. For example, using the customer's stock-in-channel position, Motorola pre-built phones in anticipation of a stock shortage well before the customer actually placed the order. Motorola often surprised the customer by calling for an order at the same time that the customer was preparing to place an order.

For the first six weeks, the Motorola team focused on fixing existing delivery problems by holding daily meetings to review every purchase order, demonstrating improved on-time delivery in order to gain trust. As replenishment improved in the short term, the relationship improved, and Motorola convinced the retailer that it could collaborate and find ways to better work together without simply building up inventory. Only then could the participants establish common goals and measures, develop formal communication points and establish peer-to-peer relationships.

Target lead times and inventory levels were established. The retailer's executive management team bought into the necessary changes to its processes, organization, and technology. Motorola and the retailer jointly created a master agreement and joint business plan that they continually reviewed and adjusted as the environment changed.

Motorola and the retailer jointly reviewed cross-linked metrics (e.g., the retailer measures forecasts and Motorola measures delivery; the retailer measures inventory and Motorola focuses on order changes). They also used the VICS CPFR handbook to determine what should be measured. Collaboratively, they used a Six Sigma approach to review the root causes of stock-outs. And together they agreed to monitor the following: product availability, order confirmation, on time delivery performance, flexibility of order quantity, sell-through, inventory, order changes, order accuracy.

To help rationalize business processes for planning, forecasting, and replenishment, Cederlund's group set up formal communication processes at multiple points along the supply chain. The processes included weekly reviews. (See Exhibit 2.) Every week during a collaborative planning meeting, Motorola and the retailer reviewed replenishment, prior/current week sell-through, inventory position, and open orders. Additionally, the participants set up a new process for continual dialog between marketing and finance. While regular telephone conferences kept the collaboration moving, the participant teams met face to face at least once every six weeks. The teams accomplished most of the operational tasks during weekly activities between face-to-face meetings. Here is a sample of their activities:

Week 1: Operations review. Review of the prior month's performance. The first 20 minutes of each meeting dealt with tactical matters; the teams reviewed any positive and negative events from the previous few weeks. The rest of the meeting focused on the strategic implications of what had happened and examined the implications for future forecasts.

Week 2: Forecasting. During the second week of each month, the teams developed their collaborative forecasts. Prior to CPFR, the cell phone retailer's forecasts were developed at the end of the week while Motorola assembled its plan early in the week. Motorola convinced the retailer to move up its planning cycle by just two or three days, which eliminated a seven-week forecast lag that had been the result of the forecasts not being incorporated until the next month's planning cycle.

Now on Monday, the retailer loads its forecast for the next month. On Tuesday, Motorola loads its forecast. On Wednesday during the weekly call, the two teams jointly resolve discrepancies line by line. The inclusion of a marketing analyst means they can immediately resolve issues related to discrepancies.

Week 3: Process improvement. Motorola and the retailer examined the issues brought up during the operations review and assigned specific actions. They also examined event timing and checked their calendars to make sure that the timing still worked. Together they reviewed the effectiveness of their baseline assumptions. Jointly they assessed the processes (e.g., whose process was working right, and what rules they should follow in the future). This enabled them to plan more effectively and to discuss replenishment policies. They also took the opportunity to review what was working and what was not working, and to discuss what was missing and what went well. Incremental changes were made over time to perfect the collaborative process.

Week 4: Financial Implications. Finance managers would attend alongside the customer account team so they could discuss the financial implications of the CPFR process. This was particularly helpful; it is easy to overlook the effect that CPFR can have on participants' period-end goals. Large public companies such as Motorola need to predict financial results within strict boundaries. CPFR allowed Motorola to increase or decrease shipments or lower its inventory. The participants could carefully plan fiscal period ends for more accurate predictions of profit and loss and balance-sheet items.

Cederlund's group then helped to redefine the organizational structure to put the spotlight on collaborative relationships. One key move: the company set up account-based operations teams. The separate functions that had previously provided input to the sales team, including customer advocacy, business operations management, finance and credit, regional marketing, product operations, pricing programs, and customer development were pulled together into a dedicated customer account team headed by the strategic account manager for that retailer. This team could directly discuss with the retailer the lead times, key time fences, operational calendars, product roadmaps, promotion plans, and key triggers.

Individual roles changed too; collaborative customer skills were now required. (See Exhibit 3.) Previously, the handset division's call-center personnel had essentially been order-takers, each handling two to three customers at a time. With CPFR as a core framework, they were retrained to become replenishment analysts focused on managing customer inventory.

After completion of the six-month pilot, the division made its big (and permanent) customer-centric move. Two new roles were added: a director of customer operations and a customer alliance manager. Order management and demand and supply planning roles were cut back and personnel were distributed into customer teams. Business planning and business analysts became part of the customer-focused teams. A single view of demand began to crystallize where once there had been multiple views in different groups within Motorola.

For its part, the retailer combined separate functions that had previously provided input to its supplier management team, including purchasing, logistics and inventory management, finance, product marketing, product operations, and sales operations. Executive support of this organizational change was critical. The retailer's new group became a supplier-focused demand planning team. Collaborative relationships were strengthened because there was direct communication on all issues between Motorola's new account sales team and the customer's new supplier management team.

Since CPFR requires a significant amount of information-sharing, it is critical to have a robust information system. Cederlund's team kicked off with a simple approach, sharing information in Excel spreadsheets. Once the business processes had managed to reduce forecast error for about six months, the team moved to a specialized collaborative system provided by Manugistics, eventually incorporating eXtensible Markup Language (XML) for direct integration with its supply chain software.

The Manugistics system then ran in parallel with Excel for about six months, from August 2002 until February 2003, at which point the use of Excel was minimized for data sharing purposes. This enabled testing of all the interfaces. While CPFR can work with a simple Excel spreadsheet, it can be error prone and takes more time to create and update than it takes to automate with Manugistics. The Manugistics system offered several other features: automated input of demand signals into Motorola's overall planning system; automated feedback to the customer of Motorola's supply commitment; automatic analysis of inventory over- or understock with signals to responsible planners; and “management by exception” instead of a line-by-line examination of inventory.

As Motorola and its retailer changed their processes, organizational structures and information systems to align with the collaborative business model, Motorola managers saw that inter-organizational relationships were changing too. New peer-to-peer relationships between participants became powerful facilitation mechanisms for transforming their respective businesses. (See Exhibit 4.)

Before the adoption of CPFR, communications took place primarily between Motorola's salesperson and the retailer's purchasing agent. If there was a problem, the spokespeople would have to relay issues to others within their own companies. The process created a lot of noise since the “primary contact” individuals often lacked all the facts and the skill set to discuss issues over a broad range of topics. As a result, relations between the participants were often tense.

But after the CPFR framework was in place, Motorola had seven or eight individuals who knew their counterparts at the retailer. The process and organizational changes facilitated formal yet trusting relationships among peers. Domain experts in marketing, finance, and supply chain in each organization were now communicating directly—and enjoying greater camaraderie. Now instead of just one Motorola salesperson going out to dinner with the retailer's buyer, as many as 20 people do so, bonding as a team and sharing information. Relationships are now based on teams, not individuals, and the quality of information shared keeps improving.

Motorola's CPFR initiative made a big difference to forecast accuracy almost immediately. MAPE (mean absolute percent error) is now a fraction of its previous levels. CPFR allowed quick reductions in inventory at the retailer's distribution center because there was less need for buffer stock. Stock-out rates are now less than a third of what they were previously.

However, Motorola sales revenue did not immediately increase since it was necessary to clear out the sales channels. Motorola reduced shipments initially while recalibrating inventory levels. It took a few months to get the inventory aligned with channel consumption. Once that happened, shipments began to pick up as out-of-stocks began to drop. The retailer soon responded with increased sales as it saw Motorola looking out for its inventory levels. The increase in shipments more than made up for the initial reduction in shipments; and established a new run-rate in sales, higher than the pre CPFR run-rate.

After the CPFR initiative had been established, the inventory that Motorola reserved for the retailer decreased by 30 percent compared to the general inventory for its other non-CPFR retailers. Collaborative forecasting and replenishment also produced a secondary benefit. Transportation costs were cut in half because Motorola was no longer shipping multiple pallets daily, many of which had been in less-than-truckload (LTL) shipments. Using its CPFR systems, Motorola could plan longer production runs, ship full containers and thus improve utilization of production capacity because supply chain set-up could be reduced—and because it was less variable.

At the same time, the results of cell phone promotions improved because the participants' marketing teams were more involved with each other and because they had greater visibility of supply chain milestones. New product launches were also executed much more smoothly. Previously, new product launches had hurt forecast accuracy involving other products due to the lack of understanding product crossover plans at the retailer.

The resulting collaboration now yields greater reliability because of the involvement of all concerned parties. With real-time data, reaction times are also faster should unexpected events occur. Prior to CPFR, retailers sometimes gave Motorola “C”, “D” and “F” ratings on metrics such as on-time delivery, ease of doing business, stock-outs, and order quality. After CPFR these same retailers gave Motorola “A” ratings on the same items.

Once shipments leveled out to match sell-through, change orders and expediting requests were minimized, and the focus on process began to uncover other opportunities such as the potential for vendor-managed inventory programs.

1) Collaboration is the foundation. The real key to a successful implementation of CPFR is the forging of cultural alliances rather than traditional supplier-customer relationships. Such alliances involve peer-to-peer relationships, common goals and measures, and governance through formal communication points. Collaborative efforts should begin by first developing internal cross-functional teams and measures and then linking them to the participant's comparable teams and measures. Linked performance measures should focus on common goals.

2) Both participants must be ready. Both participants must be focused on deeper supply chain visibility such as data on sales to the end consumer. Participant readiness must be assessed to determine the appropriate business model. A standardized alliance model might not work with every customer. No single business process fits all customers; thus, collaborative processes must be customized.

Assessments of customer readiness should gauge management stability, high-level ownership of distribution and the current or potential share of the manufacturer's sales going to each customer. It is also important to scrutinize the retailer's corporate vision and mission, its leadership position in the industry and its willingness to participate, the degree to which its distribution centers and logistics organization are centralized, the type of participation (technology or channel), and to what extent the customer uses EDI or other electronic connectivity.

Customer service focus is critical to success. A senior vice president of sales at Motorola puts it this way: “Customer intimacy is a real strength of our sales team. We've learned to adapt to how the customer wants to do business—regionally or locally. Our philosophy is simple: Never say no to an important customer. That doesn't mean we give them everything they ask for. But we always find a way to say yes to something that's beneficial for both Motorola and the customer.”

3) Time-based decision-making is key. Both participants must adopt a common calendar for planning. In addition, Motorola standardized the decision-making dates, ensuring that both participants had appropriate information and plans ready for concurrent decision-making. Now Motorola and its retailer can jointly develop broad plans for the next five quarters, and standardize on monthly forecasts for the next 12 months and weekly forecasts for the next 13 weeks.

4) Success requires formal communication points and formal processes. To ensure timely and appropriate communication, weekly formal meetings should be established. Formal meetings mean that individuals from each company have specific roles and responsibilities for process, materials, content, and accuracy. A clear escalation path is in place at each participant to send any big problems up the formal chain of command. This ensures that the appropriate people are always involved. When communication is made formal, decisions are made jointly and promptly.

5) A phased approach to implementation works best. It is important to keep the process and changes simple to start with. This mitigates the risk of change and allows time for learning. Maturity can be developed over time. Implementation can begin with any one of the collaborative processes: planning, forecasting or replenishment. An early focus on replenishment can often lead to quick wins which help to garner enthusiasm for moving forward.

6) Common goals and measures are essential. While common goals are established as part of the business strategy, it is important to operationalize these goals with specific measures that are continually assessed. Value is in the eye of the beholder, so participants must agree on metrics. To start developing these metrics, begin by sharing your own internal metrics.

7) Collaboration is focused on building relationships. The strength of these relationships determines how long collaboration will continue and how successful it will be. One of the most significant determinants will be the development of peer-to-peer relationships.

8) Organizational roles must change. The organization must change to become customer-focused rather than product-focused. Customer- and supplier-focused teams help develop the strong peer-to-peer relationships that form the basis of all interactions.

9) Technology tools, if aligned properly, are necessary but not sufficient. Motorola began with a simple Excel tool to share forecast information. While usage of the Manugistics system was more efficient, it was not necessary in order to begin to benefit from the CPFR process. The improvements were generated in the first six months of the implementation before the Manugistics system was up and running.

For initiatives as complex and long-term as CPFR, we recommend enlisting strong and educated executive support. The assumptions that matched the old way of business have to change. Implementation of a CPFR process takes time, and if it is to be done right, it requires that executives have the stomach for substantial changes in organizational roles and relationships. Sales and finance executives may be less willing to stay the course with CPFR if sales drop while inventory in the channels is cleared out. The initiative has to be driven by a passionate advocate who understands the process and the long-term value of CPFR, and that person has to stick with the initiative for at least eighteen months to make it work. But when two participants make a commitment to jointly put CPFR into action, the results are impressive. The case of Motorola's CPFR initiative serves as a compelling example of a supplier's commitment to the industry and its customers.