Collaborative planning in the supply chain – CPFR (Part 2)

By Diego Souza placeholder image.

Onboarding

For companies that want to start a CPFR pilot in their supply chain and achieve the goals promised by the process, some important tips should be considered.

1) Assessment phase of the company’s current state

According to roadmap of the process published by VICS in 1999, as a first step to be taken before implementing the CPFR, an analysis of the company's needs, values, culture, strategies, relationships with partners and history of implementing best practices should be carried out. The output of this analysis will be an indication of the areas in which changes will be necessary so that the CPFR can be implemented and the company can achieve the promised gains. At this stage, two activities stand out: development of CPFR's vision in the company and the selection of potential partners.

a) Development of the CPFR vision in your company:

Before starting the process, it is necessary that it is documented and clear to all involved what are the main plans and objectives of the CPFR in the company. A vision of the process needs to be developed. A good vision will be able to answer the following questions:

• • What are your company's objectives with the CPFR?
  • Which areas of the company will be impacted?
  • How will the process be measured?
  • What will your scope be?
  • Which partners will be involved in the process?
  • What products will be considered in the process?
  • Which stores and distribution centers will be included?
  • What is the level of technological sophistication of your company and of the chosen partners?

a) Development of the CPFR vision in your company:

The definition of commercial partners is a crucial step in the process, contributing significantly to the success or failure of the implementation.

The following issues should be taken into account at this stage:

• • Is your relationship with your partner characterized by openness and trust?
  • Do you have complementary strengths and weaknesses? For example: One is strong technologically while another is strong in market knowledge.
  • Does your partner have the necessary commitment and resources for CPFR to be a success?

2) Pre-pilot:

At this stage, companies must establish the processes that will be followed in the collaborative planning, define products and units that will be included. The indicators that will be used to measure the performance of the process must be chosen.

An essential tip is to start simple. Ideally, the steps should be taken little by little and the process should only start with information from the central DC, without including the stores, and selecting more stable and less promotional items. Another important tip is to select items that have a good relationship: low complexity X greater return opportunities.

At this stage, scopes and responsibilities will also be defined, it is important that the role of each employee is well defined in the process.

Another key item in this phase is the standardization of data that will supply the system, the language has to be the same throughout the chain, product names and encodings have to be aligned between partners so that inconsistencies do not arise.

3) Pilot execution and expansion:

The next step will be the execution of the pilot run. At this stage, the suggestion is that the VICS model mentioned above be used as a basis.

After the success of the pilot, if the company wants to expand the CPFR, the direction is to first seek partners with profiles close to those of the pilot and to increase the number of SKUs in the central DC before expanding to other DCs or stores.

Indicators

In order to guarantee the continuous improvement of the process and to maintain the incentive for the participation of the partners, it is essential to create and maintain a panel of indicators. We can divide the indicators into four broad categories: Planning Accuracy, Customer Service, Supply Chain Performance and Process Quality. Each chain must define the best way to measure each of these dimensions in a way that all partners understand. In this section we will give some examples of indicators that can be used in each of the categories.

• Planning accuracy:
  • MAP: Mean absolute percentage error. It measures the deviation from planned to actual, and can be applied to both the sales plan and the order plan. It can also be weighted by some important factor, such as total contribution margin or revenue.
  • MEP, or Bias: It measures the bias of the plan, indicating whether on average we are being more optimistic or pessimistic.
  • Suggested Order Acceptance: Generally speaking, the process will have some automatic way of suggesting a purchasing plan (sell in) based on the sales plan (sell out), which will depend mainly on the inventory policy and replenishment logic agreed between the parties. The frequent manual adjustment of these suggestions makes the process more expensive, and may signal the need to review the calculation parameters. This indicator can be measured, for example, via MAPE, or by percentage of plane lines with manual adjustment.
• Customer service:
  • OTIF: A classic service level indicator, the “On Time, In Full” is the measure of the perfect order, which arrived at the right time, and in the right quantity.
  • Out of Stock, or Break: Lost sales due to out-of-stocks can be quantified by the margin lost for each order that has been canceled due to out-of-stocks, or by the average sales of each day with zero or close to zero inventory. It can be measured at each link, reaching the retail shelf (on-shelf availability).
  • backorder: Orders that encounter a stockout are not always cancelled. They can accumulate in the management system until they are attended to or forgotten. It is important to monitor this queue, measuring it either by number of
  • Sales increasing: having the right product, in the right place and at the right time increases the chances of selling. Therefore, it is expected that the entire CPFR process will positively influence the volume sold.
• Supply Chain Performance:
  • Stock Levels: Inventory can be measured in total units, total capital invested, or coverage in days. It is also possible to measure it by comparing it to the inventory policy established between the parties, as a percentage of products outside the policy. This indicator, along with sales losses due to stockouts, are the most obvious indications that the process is paying off.
  • lead time: the reduction of uncertainty and its better management helps to reduce the average lead time between the links, consequently reducing the need for inventories.
  • Percentage of emergency transport: Air freight and unconsolidated cargo are good emergency options for unforeseen circumstances, but they can significantly increase logistical costs and must be closely monitored.
• Process quality:
  • Information quality index: The basic information of the process, mainly Sell Out and stocks, must be judged according to the following dimensions:
    • completeness: all the necessary data is in the base?
    • Punctuality: Is the data arriving on schedule?
    • Shelf life: Are sales volumes and inventory levels within normal limits? Is there any decrease in inventory without an associated sale?
    • accuracy: Does the data represent reality? Does the inventory in the system match the actual inventory?
    • Consistency: are the data the same between systems, such as the CPFR tool and the companies' management system?
    • Integrity: Are all the products sold in the product register? All outlets too?
  • Inventory Accuracy: Simpler than quality score, inventory accuracy is usually already measured in most companies. Mismatches between actual inventory and inventory in the management system can cause stockouts and overorders.
  • Compliance with the schedule: As mentioned in the challenges section, CPFR influences a number of other internal processes at each link in the chain, and involves many people. Without strict respect for the schedule, the process cannot work.
  • ROI: Implementing the CPFR can mean significant investments in staff, consulting and tools. It is important to establish a base case for key indicators such as lost sales and inventory levels, and measure the gain over time to ensure that the effort invested is worth the return.

For some of these indicators, it may make sense to measure them separately by product category, ABC rating, lifecycle stage (launch, maturity and retirement) and promotion.

Implementation Cases

Metro AG and Procter & Gamble

Metro AG is Germany's largest wholesale and retail group, operating in 2250 points of sale in 24 countries. In 2001, a CPFR pilot was started between the group and Procter & Gamble, which ranks among one of the main manufacturers of consumer products in the world.

In the pre-pilot phase, at the end of 2000, the cross-functional team that would work on the process was selected, with members from both companies. The focus of the process was defined and it would be centered on the planning of promotional products. Indicators for measuring the process were agreed, as well as the data standard and frequency of information sharing.

The pilot started the following year included 27 products from a given line, 53 stores and 1 distribution center. At a later date, the scope was extended to two product lines, resulting in a total of 130 SKUs.

The main objectives of the process involved improving the promotion planning process, reducing outages, optimizing stocks, encouraging real collaboration between partners and, finally, increasing the level of service to the end customer.

On a weekly basis, Metro and P&G collaborated on promotions and sales plans (Sell Out) for promoted products. P&G made a plan for the next 12 weeks while Metro collaborated on the next 8 weeks. If the difference between the plans of the two partners exceeded a pre-agreed limit of variation, the software used generated exception notifications. The generated exceptions were resolved with the help of historical data and known facts. The assumptions used were registered in the system and a plan agreed between the two parties was reached. Based on the sales plan, store orders were updated.

The tool received daily data from points of sale throughout each promotion to check the performance of products participating in the process. This monitoring helped detect possible problems and allowed the chain reaction to solve them.

Indicators used by Metro and P&G

The indicators used in this process were: Out of Stock measured at the distribution center, On Shelf Availability, coverage of stocks in the distribution center, coverage of stocks in stores, Order Accuracy (conversion of the sales plan into an order plan x order placed).

The result of the CPFR pilot implementation was quite satisfactory for both parties. The 46% improvement in forecasting orders for the products and stores included in the pilot and the reduction of stock levels in the warehouse for up to two weeks were some of the gains obtained with the process. The pilot's success was crucial for CPFR's expansion, which began in 2002. Metro expanded the process to 3 more suppliers in the first quarter of the year and to another 6 in the second quarter.

LCBO and suppliers

The LCBO (Liquor Control Board of Ontario), an alcoholic beverage retailer operating in Canada, started a CPFR pilot in 2003. The company had implemented new commercial strategies, which included changes in store layout, innovative displays and merchandising, and a more dynamic product mix. The changes have elevated the consumer's shopping experience, which has reacted positively. The success of these commercial strategies was detrimental to the supply chain due to increased demand uncertainty and supply failures. Furthermore, his chain was becoming more and more complex. Consumer preferences were changing, the product portfolio was expanding, their lead times were long and their suppliers belonged to the 5 continents.

The solution to the supply chain impasse was to try to improve the flow of products through a network of partners, collaborating effectively and focusing on the end user. Research was carried out on best industry practices and the LCBO's solution was to implement a CPFR process along the lines of the model suggested by VICS.

The CPFR process was customized for the LCBO, fully integrated and automated. Suppliers were an important source of inputs during the development phase. In 2003, the pilot started with 6 suppliers and in 2005 it was expanded to 21 suppliers, representing 34% of the business.

Evolution of forecast accuracy in LCBO

Sales, orders and inventory information were transmitted each week, increasing the chain's visibility. Promotion plans were jointly developed for 18 months. Sales plans were carried out by the LCBO and the suppliers, recording the assumptions used, exceptions were handled and then a consensus was reached between the plans. Ordering plans were developed for the next 52 weeks.

The performance evaluation of the suppliers was carried out quarterly, through reports with the results of each supplier and a table of results identifying benchmarks. This action was supported by suppliers, who saw it as a benefit, as it encouraged improvements and identified possible problems. Some of the indicators used in the measurements were: Forecast Accuracy to WMAPE (Weighted Mean Absolute Percentage Error), Forecast Value Added, Tracking Signal, In Stock Position, On Time Delivery e Order Fill Rate.

Evolution of inventory turnover at LCBO

The biggest barriers encountered by the LCBO in the process were the lack of trust between the players, commitment, cost and complexity of the process. Despite the challenges encountered, the results were positive. the indicator Forecast Accuracy improved by around 3 percentage points and inventory turnover doubled its result after the pilot. In addition to these, other indicators such as In Stock Position, Order Fill Rate e On-Time Delivery improved and, most importantly, consumer satisfaction has also evolved.

Implementation Cases

Many authors question the feasibility of the CPFR due to the fact that it is a lot of work, faces very high cultural barriers, and depends heavily on the accuracy of retail planning in order to function. However, the various documented implementation cases indicate that there is an enormous opportunity to increase sales and gain efficiency in the chain. For obvious reasons, cases of failure are not heralded with the same fervor, so it would be difficult for us to obtain a complete database from which we could infer the percentage of successes and failures.

Here in Brazil, 60% of the large companies interviewed in a recent Supply Chain Planning Panorama published by ILOS say they do not have CPFR, and 15% already have a pilot implemented, but do not have the process as part of the strategy to improve services and reduce costs. Only 12% of the interviewees identified the CPFR as a strategic process. Chains such as medicine, which have an abundance of available information and very high inventory costs and lost sales, are still timid in their initiatives in this regard. However, we have already begun to see some players put collaboration as a fundamental success factor to improve the efficiency and effectiveness of the chain.

Factors such as the growth of cloud computing and tools in the SaaS model (Software as a Service), as well as the greater availability of point-of-sale data associated with Big Data technologies, make the current environment conducive to a new wave of deployments from CPFR. We hope that, this time, the technological differential will be the last necessary push for this process to finally gain scale.

References

• ANDRASKI, Joe. The New Model of CPFR. Available in:https://www.bisg.org/docs/VICS-CPFR.pdf>
• AROZO, Rodrigo. CPFR – COLLABORATIVE PLANNING: THE SEARCH FOR REDUCING COSTS AND INCREASING THE LEVEL OF SERVICE. Available in:https://ilos.com.br/cpfr-planejamento-colaborativo-busca-da-reducao-de-custos-e-aumento-do-nivel-de-servico/>
• AROZO, Rodrigo. CPFR AS A SUPPLY CHAIN ​​INTEGRATION MECHANISM. Available inhttps://ilos.com.br/o-cpfr-como-mecanismo-de-integracao-da-cadeia-de-suprimentos/>
• BORGES, Wagner Luis. An analysis of the implementation of CPFR in retail companies. São Carlos, 2011. Master's Dissertation, School of Engineering of São Carlos.
• CECERE, Lora. Let's face it. We have not done a good job on CPFR or VMI. Available in: http://www.supplychainshaman.com/new-technologies/lets-face-it-we-have-not-done-a-good-job-on-cpfr-of-vmi/>.
• ECR Europe. European CPFR Insights. 2002. Available inhttp://ecr-all.org/wp-content/uploads/pub_2002_cpfr_european_insights.pdf>
• ILOS. Panorama ILOS – Supply Chain Planning. 2015.
• MILLER, Craig. CPFR Case Study: Liquor Control Board of Ontario. 2005.
• RITTER, Fernando Gomes. Proposal for implementing CPFR in an adhesive company and its customers. Available in:http://www.lume.ufrgs.br/bitstream/handle/10183/2410/000369270.pdf?sequence=1>
• SHEFFI, Yossi. The value of CPFR. RIRL Conference Proceedings, October 13, 2002.SHERMAN, Richard J.. Why has CPFR failed to scale? Available in:http://www.supplychainquarterly.com/archives/200702/>
• SULESKI, Janet. Beyond CPFR: Retail Collaboration Comes of Age. AMR Research Report, April 1, 2001.
• VICS. Roadmap to CPFR: The Case Studies. 1999. Available at:http://www.gs1us.org/DesktopModules/Bring2mind/DMX/Download.aspx?Command=Core_Download&EntryId=632&PortalId=0&TabId=785>