STRUCTURAL ANALYSIS OF THE DEMAND PLANNING PROCESS

This article seeks to present the demand planning process in a holistic way, reflecting on the factors that make it possible to properly analyze the market and the company's internal capabilities and, consequently, lead to better decisions in meeting demand. For this, we will focus on the fundamental definitions and structure of the process, which prepare and facilitate the sequential steps of analysis of past sales, current and future commercial scenario, operational capabilities, financial result and marketing and strategic vision, fundamental for a planning efficient.

DEMAND PLANNING PROCESS

Demand planning is responsible for coordinating the flows of information and physical products in a company, having relevant impacts on marketing and commercial management, on the supply of products and raw materials, on production scheduling and control, and on logistical operations. . Therefore, it influences the level of service perceived by customers and the main cost centers of a company. Planning demand efficiently is much more than simply performing an accurate sales forecast. It is, rather, correctly allocating the resources available in the company to maximize financial results, simultaneously considering market scenarios and long-term strategy.

An increasingly complex environment, with an increasing number of items to plan, competition and geographic service coverage, combined with the growth of the Brazilian consumer market, which makes planning errors more costly, have led companies to seek, in recent years, improve its demand planning mechanisms. However, despite the great efforts and investments to adopt sophisticated tools, such as Advanced Planning Systems (APS), and/or to create Sales and Operations Planning (S&OP), many times the intended results are not reached simply because the importance of some fundamental definitions for the correct realization of the monthly planning cycle is not recognized. Some of these definitions are made on a case-by-case basis, without adequate analysis and reflection, and some stages of the process are neglected or simply suppressed.

To result in better decisions in meeting demand, the planning process must be preceded by two fundamental definitions that will serve to guide the following steps: the level of grouping of information and the planning horizon. In addition, the planning process must consist of stages that allow taking into account past demand, the present and future commercial scenario and the restrictions on supply, production and distribution capacity, in order to resolve potential conflicts based on the analysis. alternatives and also the marketing vision and the company's global strategy. Finally, it is important to identify the reasons for planning errors and deviations, whether caused by commercial or operational reasons, enabling the improvement of decisions in the following cycles.

Figure 1 - Demand planning structure
Source: ILOS

INFORMATION GROUPING LEVEL

Defining how the information will be analyzed throughout the process is an important step for the success of demand planning. There are three dimensions to information organization that need to be considered: time, product, and location. The organizational areas of companies, in fulfilling their functions, work with demand information at different grouping levels. For example, while the production manager thinks about the weekly production order (time) of the MRP system per item (product) in his factory (location), the commercial manager is concerned with reaching the monthly sales quota (time) of his portfolio (product) for its operating region (location). The marketing manager, in turn, analyzes the quarterly sales (time) of the family of products under his responsibility (product) for the company as a whole (location), while the transportation manager is scheduling the daily deliveries (time) of the products. orders (product) for each customer from a Distribution Center (location). That is, it is as if each area spoke a different language, which makes it difficult to analyze information and make decisions. It is therefore necessary to choose a single language for the planning process and that this can later be “translated” for making specific decisions in each area.

In summary, the company needs to choose a level of grouping of information that is between the maximum fractionation of these, which would be the daily sales (time) of each item (product) by delivery point (location), and the extreme consolidation, represented by annual sales (time) of all items (product) of the company (location). In general terms, the greater the fragmentation of information, the greater its usefulness for decision-making, but the worse the predictive capacity. For example, it is easier to correctly plan how much coca-cola will be sold in a month in Brazil than how much cans of coca-cola will be sold next Monday in São Paulo, although the latter information is more useful for operational decision-making to meet the demand.

Figure 2 - Information grouping level
Source: ILOS

Furthermore, the more disaggregated the information is, the more time will be required for the analyses, that is, the greater the number of man-hours required to carry out the planning process, making it more expensive. Thus, it is essential to group information in such a way that it is possible to make reasonably reliable forecasts, analyze them in operational, tactical and strategic plans and with a compatible investment of time. It is interesting to note that planning with very aggregated or very disaggregated information leads to the same problems of meeting demand, but in different ways. While the ambition to carry out planning at a high fractional level leads to service problems due to the difficulty of forecasting, aggregate planning leads to errors due to the difficulty in breaking down information for operational and product mix decisions. A good practice for organizing information, which works well for a large number of companies, is to work at the product family level (similar consumption characteristics – Ex.: “Omo family”, “Coca-Cola family”), with sales monthly per installation serving the customer (eg distribution centers, factory warehouses).

PLANNING HORIZON

Another very important preliminary definition is the planning horizon that will be analyzed. Once again, the functional areas of a company deal with restrictions and characteristics that oblige them to anticipate their actions in a certain period of time. For example: the time for importing some items by the supply area can be more than three months, which obliges the area manager to plan the imported volumes well in advance. The production area, in turn, schedules production on a weekly basis, but as it has a very diverse mix, it is unable to produce all items every week, being forced to do a closed planning for the month. The transport manager, on the other hand, does not need to plan the operation so far in advance, as delivery schedules are daily. Thus, planning the following month (M+1) may be enough for the production and logistics area, but it does not allow the supply area to react to an abrupt change in the volumes already purchased.

It is therefore necessary to recognize that the definition of “for which period in the future the company will make efforts to plan” cannot be casuistic and must consider the criticality of the supply, production and distribution processes, and also the difficulty of anticipating the needs from the market. For this choice, the company must consider that the shorter the planning horizon, the greater the predictive capacity, but the more expensive and flexible its service resources must be. Taking advantage of the example in the previous paragraph, if the company planned only the M+1 period, it would need to have more stock of inputs to be able to deal with demand variability or have higher costs with emergency import orders and intensive use of air transport, a since with the regular process of 3 months for imports, there would be no time to react to changes in the market. On the other hand, if she planned for M+3, giving the supply area reaction time, her planning error would be greater. Companies must therefore, in order to decide what their planning horizon will be, balance the cost of the error and the additional costs of shortening this horizon.

Figure 3 - Planning horizon
Source: ILOS

Although each company has its own operational characteristics, a good practice in defining the planning horizon is to make a statistical forecast for the next 12 months and then concentrate the planning efforts on the first three months. The statistical forecast for the next 12 months is important for companies that deal with intense seasonality, as it allows anticipating the need for resources in periods of peak demand, properly preparing buffer stocks and hiring logistics service providers. The planning effort takes place, however, for the next three months. For M+3, it is enough for the commercial and marketing areas to signal if there will be any significant change in the market that requires an immediate reaction. There is no need to discuss the numbers in depth. For M+2, the reaction capacity of some areas, especially imported supplies, is already starting to be limited and, therefore, it is necessary to analyze the numbers in an attempt to avoid unnecessary costs later. The next month, M+1, is the so-called frozen period, in which the decisions taken in the current month (M) will be executed. Planning and analysis efforts will be concentrated there, as inadequacy in the decisions taken means loss of commercial opportunities and/or leads to the need for rescheduling planning, often with significant costs.

PROCESS STEPS

Once the information grouping level and the planning horizon are defined, the company is ready to start its monthly demand planning cycle.

Step 1 - Analysis of Past Demand (Baseline Forecasting)

The first stage of the monthly cycle of demand planning is to make a sales forecast based on the analysis of historical series. Briefly, sales forecasting with the objective of projecting the past behavior of demand into the future consists of applying an extrapolative mathematical model, which can range from a simple arithmetic average to even very complex models, such as genetic algorithms or neural networks.

This step is very important, as it allows the regular characteristics of the sales series, such as level, trend, economic cycle and seasonality, to be understood and replicated. In addition, the use of quantitative models, whether in specific tools or electronic spreadsheets, allows for an increase in the agility and accuracy of this activity. Considering that the company may have hundreds or thousands of sales series to predict, without the use of quantitative models the time that analysts would take to understand the behavior pattern of each series with the same degree of accuracy (in most cases it would be less !) would make the process unfeasible.

Figure 4 – ILOS survey on sales forecasting in large Brazilian companies
Source: ILOS

Some companies are frustrated when, a few months after purchasing a sophisticated sales forecasting tool, they notice that the model output starts to show a very high error. This occurs because sales forecasting models are incapable of understanding the entire dynamics of the market, with its promotions, product launches, media actions, among others. It is therefore necessary to keep the series of sales “clean” or in “baseline”, excluding the effects of promotions, product launches or any other commercial factor, the company and the competition, so that the mathematical models can continue to recognize the “ pattern” of the series and replicating them correctly. Delivering, therefore, good base numbers for the rest of the planning process. This is truly the job of the sales forecasting analyst.

Step 2 – Commercial Scenarios (Unrestricted Sales Plan)

In the first stage, what we did was simply use the past to try to anticipate the market's needs in the future. Despite the past giving us clues of what may happen, this is insufficient, as we know that the future is almost always different and unexpected, especially with the complexity of the current market. Thus, we need a reflection effort to try to define what will be different and consider this in our planning.

The second stage of the demand planning process consists precisely in incorporating into the “cold” number of the mathematical forecast, which reflects the past, all the commercial and market information that could influence the behavior of demand in the future period that is being planned. Promotions, product launches, new competitors, new markets, participation in public tenders, media campaigns and any other actions that may influence demand must be analyzed to “warm up” the predicted base numbers.

Figure 5 – Incorporating commercial and market information
Source: ILOS

For these analyzes to be carried out successfully, it is essential that the company, in addition to having an experienced and committed commercial team, record all its commercial and market actions, as well as their respective impacts on demand. For example: “a price 10% lower than the average was practiced and this resulted in a 15% growth in sales for the period”. These records ensure that the information is available in the company, and not just "in the heads" of the people who participated in these actions, allowing to improve decisions in future planning cycles, maintain the "baseline" of the historical series and also use causal models. to estimate more precisely the impacts of the company's commercial and marketing actions.

Step 3 - Capacity Constraints (Operational Plan)

After analyzing the scenarios by the commercial and marketing areas, the Unrestricted Sales Plan is obtained, that is, a plan that considered only the market without checking the existing capacity to execute it. As additional costs with overtime, emergency transport, fractional purchases or any actions to momentarily increase capacity can be higher than the margin obtained with sales made above installed capacity, companies must carefully analyze all operational restrictions that required for carrying out the plans.

Figure 6 – Analysis of service restrictions
Source: ILOS

To increase the agility of this step, it must be carried out separately by the company's supply, production and logistics areas, which are responsible for the resources used to meet demand. As, in general, there is not enough time for a detailed analysis, it is suggested that the planning area transform the unrestricted sales plan into inventory needs and, based on these numbers, the areas assess whether they will be able to meet these needs . It is essential that, based on this analysis, a realistic plan is set up and the critical points for its fulfillment are pointed out. That is, it should be highlighted what will not be possible to carry out the unrestricted sales plan, what can be done with additional costs and which products could be made available in greater quantities, but do not have expected demand. These are precisely the points that should be aligned between the areas.

Step 4 - Reconciliation of Plans (Financial Analysis)

The Unrestricted Sales Plan and the Capacity Plan developed in the previous steps need to be aligned, as they were built independently by areas that have potentially conflicting indicators. This alignment generally takes place in one or more meetings, in which the managers and analysts of the functional areas involved take part, to discuss, ideally based on the financial analysis of the alternatives, what should be done in each of the friction points. For example: the unrestricted sales plan indicated the need for 300.000 items of a certain product family in the São Paulo distribution center, but the capacity plan indicated that it will not be possible to make more than 280.000 units available, unless the difference be transported on an emergency basis by plane. The discussion in this hypothetical case should be based on the analysis of the profit margin obtained with the sale of these additional 20.000 units and the difference in cost between sea and air freight to bring them.

Figure 7 - Conciliation of plans (Financial analysis)
Source: ILOS

These meetings need to be previously structured by the leader of the planning process, who will identify the conflicting points in the plans and, together with a financial analyst from the company, will calculate the value of the alternatives. Without this prior work, the alignment meeting runs the risk of resulting in disagreement between the areas, which start to protect their indicators with the omission or falsification of information throughout the process.

Step 5 – Holistic View for Better Decisions (Marketing and Business Strategy)

There is no requirement that plans be perfectly aligned in the previous step, as financial analysis is often inconclusive about what is best to be done in a situation. This occurs because it considers only measurable financial values, but not the intangibles related to the company's long-term strategy. For example, one comes to the conclusion that it is not worth it, following the previous analysis, to transport an additional 20.000 units on an emergency basis, as the difference in the freight value is greater than the margin that would be obtained with the sale of these units. However, the commercial area argues that these additional units would be used to win an important contract and prevent a strong competitor from entering the market. Who can, or should, decide on this? The commercial area, which will certainly benefit from this additional sale? Or the supply area that will lose part of its result with this emergency hiring? Probably, if we leave the decisions to be taken freely by the areas, the sale will occur and the customer will not receive the product within the promised period, becoming dissatisfied with the company.

Figure 8 - Conciliation of plans (Strategic analysis)
Source: ILOS

Therefore, an additional step in the planning process is needed to mediate and decide, from a long-term strategic perspective, what is best to do in these points of conflict that could not be resolved in the previous step. For this, no one better than the company's top management to carry out this analysis and have their decision respected in the execution of the planning. This stage consists of an executive meeting, in which the company's president, who is the natural guardian of the strategy, and the directors and managers of the functional areas participating in the previous stages must participate, for analysis and decision-making regarding conflicts. As the area's indicators are affected by these decisions, the company's top executives must understand these impacts and, eventually, make corrections to the indicators. Following the previous example, the president of the company could agree with the commercial director's point of view and decide that at that moment it was important for the company's strategy to “buy” market share and prevent competitor entry. As this action would benefit the commercial area indicator, it could define that the additional cost of freight would be debited from the commercial cost center, instead of putting it as an additional cost of the supply area. The best decision for the company would have been made in the fairest way for everyone.

ANALYSIS OF ERRORS (CONTINUOUS IMPROVEMENT)

Like any process, it is expected that demand planning can be improved from the detection of nonconformities, or errors, between what was planned and what was actually done. Error analysis allows the company to make the necessary corrections and make more appropriate decisions for the next planning cycle. Errors can be the result of operational failures, such as unforeseen equipment breakdowns or delays, or commercial failures, such as errors in the sales plan.

Measuring and identifying the causes of errors should therefore be one of the main objectives of those responsible for the planning process. To measure the commercial planning error, the company can make use of indicators such as the mean absolute percentage error (APE) and the mean percentage error (MPE), calculated with the closed planning numbers and the sales made (and billed) in the period. Identifying its causes is not so simple and will require constant observation of market conditions. The participation of the commercial area for this is very important. Measuring operational deviations is not common practice in most companies, but it needs to be practiced. All equipment breakdowns, emergency maintenance, delays should be measured in hours and transformed into product numbers. For example, in one hour this machine is capable of producing 500 items. If it was down for 3 hours due to a breakdown and repair, a capacity of 1.500 items was lost. Thus, the company would be able to estimate how much of the demand was not met due to operational problems and how much was due to commercial planning errors. This systematic measurement allows the company to identify potential problems in operation and correct them.

In general, these indicators are presented and discussed in the plan conciliation and executive meetings. This allows the critical points for the realization of future plans to be identified and highlighted, favoring the execution of these plans.

CONCLUSIONS

This article sought to present the demand planning process in a holistic way, highlighting some good business practices to help executives involved in this activity to improve the preparation and execution of the steps that lead to better decisions, reducing costs and increasing customer satisfaction.

BIBLIOGRAPHIC REFERENCES

Wanke, P.; Julianelli, L. Sales Forecast: Organizational Processes & Qualitative and Quantitative Methods, Publisher Atlas: Rio de Janeiro

ILOS – Institute of Logistics and Supply Chain, Research: Analysis of the Demand Planning Process and S&OP in Brazilian Companies, 2009/2010.