Boldly Overhauling a Broken S&OP Process

It is a story of transformation and advancement at a $1 billion biotechnology division of a leading corporate materials science innovator.

This company had more than 10,000 stock-keeping units (SKUs); 16 manufacturing sites in 7 global regions; hundreds of product lines, product groups, and business segments; 3 key distribution centers on different continents; more than 50,000 individual data points to forecast throughout a 24-month horizon; and multiple separate forecast models that did not connect or share a common format. The challenges were significant.

Company leaders recognized that their S&OP process was broken. Procedures were not codified; conversations broke down; and when meetings adjourned, everyone returned to private silos to complete their work based on intuition and past historical performance alone. Adding to the issues at hand were unsynchronized enterprise resources planning (ERP) platforms.

In addition, the sales forecasting system did not have a report module, was outdated, and was unsupported from a technical perspective. The demand-planning department relied on a single, third-party contractor who knew the obsolete programming language. There was no budget for a new system or for additional demand planners. A bottom-up forecasting approach at the SKU-by-region level was used, and no one reviewed what the top of the pyramid looked like in dollars.

Manual processes had been unchanged for several years. Huge Excel worksheets and Access databases were updated regularly to interface with the demand-planning system. A great deal of time was spent at the beginning of each month creating forecast error reports, but no one analyzed the results; there was no time. It was all about the monthly process.

The phrase “That’s the way it’s always been done” was a common response to queries. Still, despite all these handicaps, it was possible to execute a new, highly successful S&OP initiative. Team members first defined the problems. Then, they established requirements for the new process, selected a concept using decision-analysis methodology, and generated a detailed design for the new state. Lastly, the new process was implemented. The following step-bystep guide kept the project moving.

Step 1. Simplify demand forecasting
The first step is always the most difficult. It involved team members using the unit-demand forecast process to create the net requirements for the weekly supply plan. These, in turn, drove manufacturing master production schedules. The accuracy of that forecast ultimately was responsible for the division’s level of on-hand inventory, revenue, and cash flow.

Before a new S&OP venture could blossom, it was essential that the inputs to the process were credible, validated, and simple:

• Quantitative processes needed to be automated. The global demand planning team required greater control of the forecasting system—not the other way around.

• Qualitative processes (such as collaborating with both internal and external customers, sales and marketing groups, and global supply chain team members) needed to be encouraged. The most significant exceptions between forecast and actual results monthover-month needed to be reviewed, confirmed, and updated as necessary. These guidelines helped S&OP participants look to the future:

• Forecast less, not more. Forecast the volume (at higher aggregate levels), and manage the mix (the lowest-level data points at the bottom of the forecasting pyramid).

• Emphasize teamwork, not formulas. Collaborate and communicate with those who understand current sales and marketing trends.

• Focus on process improvement. Reduce forecast error and eliminate bias by streamlining and taking control of the monthly process.

• Simplify and reduce multiple time-consuming portions of the current monthly process to improve the team’s ability to forecast more accurately.

• Add value by reducing the number of hours spent crunching numbers, and invest more time in continuous improvement applications.

Step 2. Reconcile units and US dollars
Experience had shown that the average selling price per SKU, when extrapolated across multiple price listings by customer or across international borders with multiple local currencies, was consistently inaccurate because of continuously shifting variances in volume across these same metrics. Here is where the finance department stepped in.

A worksheet was updated monthly to record net rebate pricing by SKU and global region. It proved very helpful to the demand-planning team. For this worksheet, it was necessary to extract the data that planners need from the monthly finance report. Also, guidance indicating the appropriate foreign exchange rates to use was required.

Over time, this process determined when the roll-up in dollars of the unit-demand forecast was unusually high or low. This led to a monthly process whereby the demand planner responsible for the monthly upload filters and sorts the data for a percentage-change “reasonableness check.”

Step 3. Assemble the project team
Here, it was important to consider the type of environment and existing organizational structure. In a smaller, simpler manufacturing situation, the standard S&OP protocol would have all key functional groups represented. But for this large company, the S&OP group included business operations, commercial operations, finance, and demand planning.

Input from an information technology business partner and a project engineer also were useful to the team. A project sponsor was recommended, and steering committee members served as vice presidents representing the primary functional groups.

Step 4. Define the problem
This was a time for some investigation. S&OP team members asked the following questions:

• What worked previously, what doesn’t work so well, and why?

• Who should be listening for the voice of the customer (VOC)?

• Where is the disconnect between suppliers and customers—as well as between inputs and outputs?

• How were multiple forecast methods used, what unit of measure was factored in, and were there common denominators to link the various forecasts?

• Who has the data and, more importantly, the right data?

• Who does the work?

• Who makes the final decisions?

Step 5. Establish requirements
Even more information was gathered by prioritizing customer wants and needs, completing a situation appraisal of the current process, and identifying actions critical to quality characteristics. VOC requirements were segmented into categories including essential items, those that improved analysis, and those that offered greater efficiency and operational capabilities. Other work included developing an accurate financial forecast in US dollars; a unit-demand forecast reconciled to US dollars; and a single, integrated process to link the two concepts.

Step 6. Select the concept and detail the process design
It was important for S&OP teams to brainstorm and review multiple potential process concepts. Meeting all of the project goals is impossible if there are multiple, misaligned forecast processes and systems. Team members asked questions such as: Do we lead with units and roll up to dollars, or do we lead with dollars and drill down to units? Do we keep parallel forecast processes in place and reconcile at the end of the month, or do we consider eliminating one of the processes by rolling up the unit forecast into dollars and using that as the financial forecast?

Eventually, a hybrid concept surfaced: The finance department would create a report template for all functional groups. Multiple worksheet tabs would connect data provided by one group to the tab for use by another functional group.

The commercial operations team would provide the global regional sales forecasts in local currencies across each of the businesses. Data cells were conditionally formatted so that, if month-over-month changes occurred within certain thresholds, a “comments” cell would be highlighted, alerting the individual responsible for the data to include a high-level commentary about the change. These comments would be visible to the other functional groups in their respective worksheet tabs.

The finance department then would provide the budgeted foreign exchange rates for the key local currencies as reported across the global regions. Although more than 50 local currencies were tabulated each month, the global sales regions settled on using the US dollar, Japanese Yen, and European Euro for reporting.

The business operations directors would review the aggregate data and provide the final financial forecast for all businesses in US dollars. This forecast would drive the function and have a direct relationship with the detailed, SKU-level unit-demand forecast.

Next, the demand-planning team would create a statistical unit-demand forecast based on the most recent historical actual sales performance by SKU by region. Net rebate pricing by SKU and region would be applied. An aggregate financial forecast would be created and compared to the business operations final financial forecast. Any deviations would be balanced appropriately across each of the businesses for a top-down forecast approach. The final unit-demand forecast would be reviewed for reasonableness and accuracy and published at the end of each month to the division’s primary ERP system.

At a monthly S&OP global demand review meeting, each of the businesses would be examined individually. Then team members would compare and contrast the assumptions made among all functional groups, review past performance, and discuss unit-demand forecast accuracy and bias metrics. For cases when consensus was not reached, people would either agree to disagree with documented commentary, settle in the middle, or plan an off-line discussion to work toward a resolution prior to the management business review meeting with the division president.

Step 7. Implement the process and keep it on track
The implementation phases for such projects can be difficult, multi-dimensional, and time-consuming. Yet in the end, all project goals had been achieved. An integrated process had been created and followed throughout the project. Tollgate review meetings took place along the way to ensure that all steering committee members were aware of the project details and fully approved the stages. It was a rigorous process, but it ensured that all key players had a stake in the game, one stage of the project at a time.

Key success metrics were implemented to measure performance for unit-demand forecast accuracy, bias, and the variance between the unit-demand forecast in US dollars as compared to the financial forecasts across the businesses. Additional comparisons were reviewed each month between the business operations and commercial operations financial forecasts.

Finally, training for all participants and stakeholders was completed, and the change-management plan for transitioning from the current state to the new process was shared. To support the shift, a proposal and business case for the necessary new tools and systems required also were presented.

The moral to the story
There are several key takeaways of this S&OP project tale. First and foremost, a substantial improvement in teamwork can lead to even greater gains in organization-wide synergy. At this biotechnology division, unit-demand forecast accuracy improved 8–15 percent, depending upon product category, sales region, and classification.

The new process also created a template that enabled all S&OP stakeholders to view data and information in a similar format. It further created cross-functional bridges and connections that hadn’t existed previously or were, at best, vague. As those connections strengthened, so too did the clarity of communications. This led to better decision-making analysis on both sides of any particular issue. It also heightened focus on the longer-term, 18–24 month planning horizon, rather than a fixation on how the company might end the current month.

S&OP is not so much about reporting methods, systems, or tools as the people who make the process. They are the lifeblood of the process, and they need trust and teamwork to succeed. S&OP is the foundation that unites these functions and enables employees to share ideas and projections using a standardized roadmap.

Perhaps most importantly, all team members and executives must admit shortcomings and have a sense of urgency to continuously adjust and improve. It takes patience and fortitude to successfully navigate a never-ending barrage of data and challenges. In the end, shake off the critics, smile, and keep moving.