Sales & Operations Planning: How to avoid the 5 key pitfalls
During the past few years, Sales & Operations Planning (also known as S&OP) has been one of the most popular buzz words in the world of supply chain management. For years, market studies have shown that companies understand the importance of this high-level planning process for the successful achievement of strategic objectives. Also in our own experience, we have felt a growing interest of various companies to implement S&OP. By this time, most supply chain professionals would agree that S&OP is a formal process in which all critical business functions agree on a common action plan to balance long-term supply and demand to reach the company’s strategic objectives. With the number of S&OP processes increasing, we have come to the next stage of process maturity: the stage to learn from our mistakes.
Based on discussions with a large number of S&OP process owners, we have found five common causes of poorly performing S&OP processes.
- No support in key functions.
- No understanding of the decision variables.
- Too short a planning horizon.
- Too much detail … or too little.
- No link with operations … or strategy.
No support in key functions
The most often heard remark amongst S&OP managers is probably: “My S&OP process is running OK but how can I convince sales / marketing / top management to support it?” In many companies, S&OP has become a supply chain process: the process is coordinated by supply chain, all meetings are prepared by supply chain people, and all activities are executed by supply chain. The only involvement of other departments is the presence of the functional managers in the appropriate meeting. Clearly, S&OP will not deliver to its full potential in this situation.
From our experience, we can testify that S&OP decisions only become reality if they are supported by all critical functions in the company. That certainly requires the active involvement of sales and production management and of the operational people responsible for the company. In some cases, it might also require research & development (if new product development is business critical), purchasing (if raw materials / components are critical), human resources or even critical suppliers and customers.
Many companies develop interesting schemes to make all departments adhere to the process. Some companies even link bonuses to S&OP adherence. In these cases, we often see that process adherence increases, but company performance does not. Although everyone adheres to the process, the process has little influence on the real operations. Very often, the real problem is not that the sales / marketing / management team does not support the process; the problem is that they don’t see the point in supporting it. The reason for that is that nobody actually knows which decisions need to be taken during the S&OP meetings.
No understanding of the decision variables.
The S&OP process in itself is well described by several sources. Enter “Sales & Operations Planning” into Google and within 10 minutes, you have a full process description explaining the different steps in the process. Read the documentation, convince your colleagues to come to the five monthly meetings, develop a few reports and you’re done. Unfortunately, the process you have put in place might not deliver any concrete results. Why? Because you missed step 1 in the implementation track: decide what you want to decide upon during the process.
Let us look at an example: a company we recently worked with realised that it could influence the demand in certain sectors by changing prices slightly and by bidding on quarterly contracts. One of the S&OP decisions became “What volume do we want to contract for the next quarter?” The S&OP team realised that to make that decision, it needed a good view on market potential and on possible product availability for the next three months. Very quickly, it concluded that the sales forecast needs to become an estimation of the market potential and that the output of the S&OP was a sales target. Simple, is it not? But how many questions would this have raised if we had started from the process instead of from the decision?
Based on well-defined decision variables it is also much easier to assure correct implementation of them in operations. Typically, S&OP processes have many outputs: consensus sales forecast, production targets, financial targets … It is important to understand that some outputs are more important than others. In a well functioning S&OP process, some outputs become strict guidelines for operational planning processes while others are more like a management outlook. The difference is strongly linked to the key decisions in the S&OP process.
Let us stick to the same example: if we consider that market demand can be influenced by the sales force, the sales objective becomes a strict guideline. If however, we consider that sales volumes are not a decision variable, it makes little sense to consider forecasts as anything more than a best guess. In the first case, one can put in place a process to follow demand and shape it to the sales objective. In the second case it is much more important to have a process to detect important deviations from the forecast and adapt operations quickly to changes in market demand.
A clear view on the decision variables finally simplifies the definition of the right performance indicators. Choosing the right performance indicators is a crucial step in the process design. We cannot underestimate the effect clear KPI’s can have on people’s behavior. By monitoring the right performance indicators – and by acting upon them – a company’s management can quickly influence the strategic direction of the organization.
Too Short Planning Horizon
Although all leading resources on Sales & Operations Planning recommend a planning horizon of 12 to 24 months, we come across many companies that do S&OP on a 1 to 3 month horizon. The most recurring explanation for this phenomenon is “our sales people cannot forecast in a reliable way what is going to happen after that”. However, the reliability of the forecast should not be the main question. The most important questions to answer in determining the appropriate S&OP time horizon are “How much time does it take us to implement the S&OP decision?” and “How long will the S&OP decision influence our operational performance?”
Let us look again at two examples. During the implementation of an S&OP process for a company in the building sector, the project team realised that the supply chain was subject to a very important seasonality with very high demands in spring and autumn and very low demands in winter and summer. Because of the company’s stocking strategy, this implied important stock build-up in some periods, and significant stock reduction in other periods. Every month, the S&OP team needed to reassess whether the inventory was at the right level to live through the next demand peak. As a result, the company needed to see at least the full seasonal cycle in every S&OP run. A 1 to 3 month horizon made no sense at all.
In another company, the project team identified that the volume commitment to some raw material suppliers was one of the key S&OP decisions. Because of the extremely high demand on these suppliers, the company needed to commit volumes on a 3 to 5 year horizon. It is evident that even a 12-month S&OP outlook is not sufficient in this – rather extreme – case but that a 5 year horizon is appropriate.
Too much detail … or too little
One of the hottest discussion points during the (re)design of an S&OP process is the level of detail. While some people strive to the highest possible detail, others push to look only at global figures. Both parties have a point. The “detailists” often argue that it is impossible to check the impact of certain constraints, such as production constraints or customer requirements on a global level. As a result, they aim for a plan on customer-article level. The “globalists” position themselves on the other side of the spectrum, arguing that S&OP should only look at the big lines. They conclude that S&OP should operate on 4 or 5 big product families.
The best practice is well-known. Leading authors recommend S&OP on ten to twenty product families. But how to identify them? Again, we suggest that you start from the decisions the company wants to make during the S&OP process. The level of detail of the S&OP process needs to be the highest possible aggregation level that allows taking the required S&OP decisions.
Sometimes, the different decisions in the S&OP process require information on levels that are “perpendicular”. In those cases, the S&OP will require a tool that is able disaggregate and re-aggregate data seamlessly so that functions looking at data from different angles remain in sync at all times.
Let us illustrate this with an example. A company in the carpet industry wanted to introduce a simple S&OP process that was based on the creation of a consensus forecast and the translation of that forecast into a high-level production plan. For the sales department, the forecast level was clear. All sales people reasoned in quantities per design. The production department however was not at all interested in the design. The production capacity was mainly influenced by the color of the carpet. In this case, there is no compromise possible. Asking sales to forecast on combinations of color and design created far too much complexity. Only a tool-based solution can solve this kind of conflict.
No link with operations… or strategy
The last pitfall has more to do with the interfaces between S&OP and other planning processes than with the process itself. Although the S&OP process in itself can be a very powerful tool to direct a company’s operations towards strategic goals, the success of the process strongly depends on the strength of the upstream and downstream interfaces. If the S&OP is not formally linked to the strategic planning exercise (such as a 5-year plan), it becomes very difficult to make it point to the right strategic goals. If, on the other side, the S&OP is not formally linked to the operational planning processes, its decisions will not be fully implemented in operations. To make a real formal link between planning processes, there are two conditions: the decision of the upstream process must become a constraint for the downstream process and the link must be implemented as automatically as possible.
An example will make this clearer: a company in the food ingredients business, the management wondered why the S&OP process had so little impact on operations. In our analysis, we found that the S&OP team spent a lot of effort defining which part of the European market would be delivered to from each of the European plants. In reality however, customer orders were allocated to plants based on the same basic rules every month. The company had no way to impose the S&OP sourcing on the order entry process. Only when the company implemented an automated interface between the S&OP results and the company’s transactional system did the S&OP process got a real grip on operations. From that moment onwards, it gained respect from the different departments involved and it became a real management tool.
Today, numerous successful implementations have proven again and again the value of a performing S&OP process. But, although best practices are described by many handbooks, the successful S&OP owner needs to avoid five dangerous pitfalls.
- S&OP processes that are not supported by all key stakeholders. Very often, the real problem is not that the sales / marketing / management team does not support the process; the problem is that they don’t see the point in supporting it.
- The S&OP team does not know what decisions it is expected to take. The design of the S&OP process becomes much easier if the desired outputs are well described. Moreover, it allows definition of the right performance indicators.
- S&OP processes with a too short a planning horizon. The planning horizon must at least be long enough to implement the decision and to oversee the impact of the decision on operations.
- S&OP processes with too much or too little detail. Look for the highest possible aggregation level that allows you to make the defined decisions. If you need different levels that are “perpendicular” to each other, implement an automatic disaggregation / re-aggregation logic.
- S&OP processes that have no link with strategic or operational planning processes. Define which strategic decisions will constrain the S&OP decisions and which S&OP decisions will constrain operational decisions. Try to implement automatic interfaces with operational planning processes.