Vendor Accent

Demand Planning: The first step in Supply Chain planning

Anand Chatterjee describes the features that must be present in a worthwhile demand planning tool.

Demand planning or sales forecasting is one of the most important aspects of any organization, be it in the services or the manufacturing sector. A services organization estimates demand for its services and thereby gears itself up to service demand. A manufacturing organization estimates demand for its manufactured goods and works towards activities such as the supply of raw materials, production capacity, distribution etc. Demand planning plays a strategic role in any organization as the planning for a lot of other activities depends on the accuracy and validity of this exercise. For example, sales and operations planning is an important function and in some organizations this planning cycle is triggered once the demand forecasting cycle is closed. There are many pieces of software available in the market which help us conduct demand planning in an effective manner. One of the most widely used of these is Microsoft Excel. Most of the ERP products like SAP, Oracle Applications and SCM products like i2 have Demand Planning functionality available in their suite of product offerings. This article explores some important functionalities and features that are useful for organizations in demand planning.

• Statistical forecasting: Most demand planning exercises start with a statistical forecast. There are various models, each catering to different behavioral patterns shown by products and markets. These include univariate models, linear models, the multivariate linear and non linear models, seasonal models, Croston’s model, mixed model etc. The list is virtually endless. They may look like small words but selecting an appropriate model for each of the products in a portfolio can be a time consuming and intricate task. There are no shortcuts here. A detailed simulation exercise needs to be carried out to select the best model for a product and market. Statistical forecasting models need to be continuously tested and refined. This means that the demand planning tool should also support a simulation environment and also the ability to compare different forecasting models. Depending on the way that data is stored in the demand planning tool, statistical forecasting can be done at various levels. There can be a top down approach or a bottom up approach. A top down approach means carrying on statistical forecasting at the highest level and then breaking it down, while the bottom up approach is the exact opposite.
• Consensus planning: The demand planning tool should support consensus planning features since demand planning is rarely the work of a single person or a single department. Demand planning is often a collaborative exercise between different departments and people, who bring in their years of expertise. That is why a tool should be able to capture their inputs on top of statistically forecasted numbers.
• Promotional planning features: The demand planning tool should also be able to handle promotional planning. An extensive promotional planning feature is a great asset for any organization. It helps plan promotions and the effect of said promotions on other products, like cannibalization. Cannibalization can be extremely difficult to capture as it not only affects one’s own product lines in a similar category but also products in other categories.
• Lifecycle management: Planning for the demand of a product spanning its lifecycle is a complex process. They may not be a simple introduction of new products or phasing out of existing products; the situation could also call for replacing an existing product with a new product or multiple products. Product substitution functionality should be an integral part of a demand planning tool. This might seem extremely simple but technically it requires a lot of features, like the ability to copy historical sales of one product into another, the ability to play around with he sales figures of one geographical area in another area etc.
• Seasonal planning: Seasonal planning is an intriguing process. It can be a difficult thing to simulate in statistics with a reasonable degree of accuracy if demand patterns are not regular. The complexity is due to the fact that festival seasons can fall in different months of the year in different years. The time span or the duration of a particular season could be different in different years. For example winter can be lengthy one year and shorter next year.
• User interface: Most organizations start demand planning with Microsoft Excel. Any organization would vouch for the fact that Excel is easy to use and over the years they have become quite comfortable with it. Thus, it makes great sense if the user interface of the demand planning tool is comfortable and user friendly. This makes it easy to get acceptance from the end users.
• Data management and archival: Another important feature of any demand planning tool is the ability to churn a huge amount of data in a reasonable period of time. It should also be able to archive old data for reference. This archival process should be easy and should not affect the current functionality of the product. If a demand planning tool is built on the data warehousing backbone it can have great ability to play around with data in many dimensions. This also makes it feasible to have statistical data forecasting at various levels not only at the lowest level at which data is captured. A data warehousing backbone also makes it easy to look at the data’s various dimensions and levels increasing the utility of demand forecasting and planning manifold.

There are various other interesting features which would be important for an organization. Demand forecasting and planning is the first step in most planning cycles in any organization. Any errors that creep into the numbers at this point have a ripple effect later on which only gets amplified. This phenomenon is popularly known as the Bullwhip effect in supply chain. With the ever changing nature of the environment that an organization is operating in along with shortening product lifecycles and other competitive pressures it is imperative to have a demand planning tool which should be able to handle the complexities of the business not only today but also for the future needs of an organization as it grows.

Anand Chatterjee is a consultant with a software multinational corporation. anandc2002@email.iimcal.ac.in