Effective Demand Planning

This white paper captures a number of key lessons from Oliver Wight’s 50+ years of experience in demand planning with some of the world’s best known organisations.

Arguably, all successful business planning and supply chain optimisation begins with effective demand planning. It is a fascinating subject and while there are always significant implications to consider in terms of geography, organisation, markets and customers, there are ten keys to demand planning which apply consistently. Here we deal with five of those principles – those that form the critical elements of a successful demand planning environment. And we explore the fundamental behaviours, processes and tools that are required to establish accountability for the demand planning process, to maximise benefit from the supporting tools, and to formalise the key inputs.

1. Ownership and accountability

An essential key to establishing accountabilities is to understand the definition of a demand plan:

“The demand plan, or forecast, is a formal request from sales and marketing to the supply chain to make the relevant materials and capacity available at the time that they anticipate the customer will require them.”

In short, the forecast is a formal request from sales & marketing to make capacity become available. Sales & marketing become accountable for the forecast; supply chain operations only have authority to make product if there is a formal request to do so from sales & marketing. This is a crucial concept and important to grasp from the outset.

Establishing this accountability calls for an Integrated Business Planning (IBP) process, and this process must contain a formal monthly review of demand.

These days Integrated Business Planning (IBP) or advanced S&OP, is at the leading edge of management thinking and practice. Originated by Oliver Wight, S&OP has developed from its production planning roots in the 1970s to the fully integrated supply chain collaboration process that Integrated Business Planning has become today. Unlike S&OP, IBP brings with it, a truly strategic perspective, and integration is what distinguishes it from its predecessor.

Led by senior management, IBP is a common sense process for aligning company plans each month, to allow organisations to most effectively allocate their critical resources – people, equipment, inventory, materials, time and money – in satisfying customers in the most profitable way. There are important rules in terms of structure for the demand review to ensure accountability.

Firstly, the review should be chaired by the sales & marketing director (if your business has separate sales and marketing directors, a choice would need to be made but experience shows responsibility falls more often than not, to the sales director). The chair signs off the 24-month volume and financial forecast – the formal request to supply – so it is crucial to have the right person in the chair. Secondly, the demand review should be facilitated by the demand manager, whose role it is to ensure all participants are fully prepared for the meeting. Attendees are drawn from marketing, sales, customer service, finance and key account (or channel) managers. Thirdly, inputs from several sources are needed to produce the long range, 24-month forecast. Marketing provides information such as brand plans, product and consumer promotions; and sales shares its customer plans and targets for new and existing customers. The ability to relate the forecast to actual business and commercial plans is critical. It also ensures that sales and marketing are truly engaged in the
process.

Finally, input from the supply chain team is also vital, because they need to have a clear picture of the overall supply chain. They need to understand the flow of material through the supply chain, enabling them to identify, for example, any points where inventory may build up with customers, which could cause a decrease in demand for future periods. Alternatively, they may have to pre-build stock to meet a seasonal spike in demand.

2. Statistical forecasting

By its very nature, statistical forecasting is data rich and requires systems and tools as an enabler to produce the required output. There are, however, some overarching advantages and disadvantages with statistical forecasting.

Advantages include:
• It is easy to generate a large amount of numbers
• The process is consistent
• For large numbers, it is impossible to do otherwise
• It gives a strong indication of what will happen and is better than a guess

The disadvantages:
• Filtered history is essential
• It cannot produce predictions based on anything other than history (knows nothing of future plans)
• It can be a dangerous tool in unskilled hands
• People must understand strengths and limitations of the model – who has access?

Statistical forecasting is very much about taking the benefits of your system functionality and using it as an input for demand planning. Let’s consider, for example, how to improve the forecasting of A-class products, versus B or C. With some businesses, it is appropriate to use statistical forecasting for C-class products, even though the forecast may be less accurate, because this frees up sales & marketing to focus on A-class products thus improving A-class forecast accuracy. This will provide a better outcome in terms of working capital.

An important key performance indicator for every demand manager should be the improvement of statistical forecast accuracy, which will result in cleaner data and sharper focus on A-class products. Forecasting tools should provide valuable outputs, which make it easier for sales & marketing to produce accurate demand plans.

3. Management of assumptions

Whatever the business sector, there are always certain facts about the market that may affect the assumptions.

Drivers, for instance, could be material or component prices. Oliver Wight worked with a steel company which supplied product to copper mines. We calculated that if the price of copper were to increase, the demand for our client’s steel product would increase by a certain percentage. The copper price caused a spike or decline in steel volume three months later – so the driver for this business was actually the price of copper over 24 months.

Levers are the factors over which the business has full control – pricing, promotional activity and new product introduction, for instance. These are in the assumptions. Businesses need to focus on the 24-month volume forecast resulting from the demand plan. To make it easier to understand the assumptions behind this, we ask ‘Under what conditions will this forecast become true?’ For example, if 3% growth is forecast over 24 months, there could be a number of
assumptions supporting that growth. Will it come from a general growth in that sector? Is it driven by a new promotional activity? Will it come from increased market share?

These assumptions are time-phased across the 24-month period – if a peak is seen in the forecast, it could be related to the increase in market share which is, in turn, related to the demand plan, account plan or whatever plan is appropriate for that business. It is vital to get this right since the integrity of assumptions is what drives the process.

An important point to remember is that behind any market group or product sector, there are usually 5-10 real assumptions that should be documented and managed over time. If the list extends to 20, the business has probably lost sight of the real facts, drivers and levers. All assumptions relate back to a process of consensus – the multiple inputs from marketing, sales, business development, strategy groups and, ultimately, the customer.

All inputs are fed to the demand plan and therefore the demand manager. The role of the demand manager is to pull together all these inputs, making sure all the assumptions have integrity, and match the baseline forecasting system to the assumptions going forward. There is no such thing as a 100% accurate forecast, but the demand manager can ensure that all assumptions support the forecast and, thus, that the forecast is realistic.

4. Demand Control is an essential part of Demand Planning

Time factors are an important part of demand planning or forecasting. Cumulative lead time is the time it would take to restore a product to stock if that product was out of stock today. In our experience we have found that across many businesses the time fence is typically situated at a point around three months in to the planning horizon.

Outside three months, demand planning is about balancing supply to match demand. Supply has enough time to react because it is outside the cumulative lead time, so if the demand plan alters, supply can react to that change.

Outside the time fence, the monthly demand review meeting can sign off the formal request and change the forecast on a monthly frequency. Inside the time fence, the frequency changes from monthly to weekly, because demand and supply must match. In other words, outside the time fence, the business is in unconstrained forecasting; inside the time fence, the business has to manage the constraint of supply.

Inside the time fence – and therefore the lead time – operations and supply chain have less time to react, so demand control is called for and the demand manager is formally responsible for the demand control process. The key elements of demand control are matching actual sales to the forecast. There are inputs from the aggregate sales plan, which is the formal request from sales & marketing; the assumptions discussed earlier in this document; the customer sales plans; the actual orders; and, if the business operates in a distributed environment, there may be additional replenishment orders for other warehouses.

The role of the demand manager is to monitor the orders and ensure they arrive in the time and volume anticipated by the demand review’s formal approved forecast. Within this there are various processes, such as forecast consumption and forecast roll.

If 200 units remain unsold at the end of a month in which 1,000 sales were forecast, the process is to ascertain whether these 200 units should be rolled into the next month or dropped. If they are dropped, that is a change to the formal request. The demand manager is responsible for communicating any changes in the formal request through to supply chain and the supply planning manager.

Again, demand control places accountability with the demand manager and sales & marketing, and it is up to sales & marketing to say whether the formal request has changed or not. Supply chain and operations should stick to the supply plan that was agreed in the monthly S&OP process until they are informed of a change to the formal request. It is important too, to understand how to manage exceptions. Managing forecast accuracy enables the demand manager to calculate the normal demand variation for a product.

The weekly planning process would highlight activity outside normal variation patterns. A weekly demand review meeting would be held, at which sales offered explanations for why certain products were outside the tolerance levels. Reasons could include customers ordering early – which would mean there is no need to change the formal request. Alternatively, something significant may have happened to affect an assumption, so a new forecast – and formal request – would be issued to supply. The demand manager and supply planning manager would look at whether the forecast was achievable and if it could be managed within the supply chain.

The key to demand control is to understand normal demand and the range of tolerance that can be accommodated within the supply chain. This should be used as an input to safety stock and inventory policies. A-class products should be the main focus and it is here that tolerances should be as tight as possible. If the tolerance for A-class products can be made tighter each week, the result will be an improvement in working capital. This, in turn, will improve forecast accuracy; forecast accuracy results in less safety stock; less safety stock results in lower working capital. That is the focus for the weekly demand planning process.

5. Measuring and managing forecast inaccuracy

At the point of cumulative lead time, we should save the forecast so we can understand and measure forecast accuracy at this point. It is likely that every product has its own cumulative lead time, but to keep things as simple as possible, apply a common cumulative lead time to a product group or family. Forecast accuracy should be measured at that time fence. If cumulative lead time is 13 weeks, save the forecast at that same stage. Therefore to check the accuracy of the plan for July, you would need to go back to the forecast entered into the system in April and measure that to determine forecast accuracy.

Building the foundation for Integrated Business Planning

These five keys to managing demand should provide valuable insight to the weekly and monthly demand planning process, and should provide a solid foundation on which to build a true capability for Integrated Business Planning.