Effective Demand Planning and Forecasting

Author:  Dr. İlker Guclu Latest update:  25 October 2026 Related:  Demand Planning & Forecasting Training , Demand Planning Consulting , What Is S&OP

In brief: Demand planning is the process of forecasting the volume and timing of demand for products or services using statistical methods, market knowledge and expert judgment. Demand and sales are not the same concept. Forecasts are the input to inventory, production and procurement decisions. This article covers the factors that affect demand, the steps of generating forecasts, forecasting methods and the corporate processes that use forecasts. To see how forecast error is measured, see How to Measure Forecast Error: Bias and Percentage Error Metrics.

What Is Demand Planning and How Do Demand and Sales Differ?

Demand Planning encompasses all planning activities a company performs to forecast the volume and timing of demand for its products or services.

Note that, customer demand and recorded sales are not the same concepts.

Demand expresses customers’ need for a product or service. Sales are the transactions that satisfy that demand.

When forecasting potential sales, using the forecasts generated by demand planning is a reliable and correct approach.

Sales data does not reflect the whole of demand. Orders that could not be fulfilled because of stockouts, customers who gave up buying because the product was not on the shelf, and deliveries postponed because of capacity constraints do not enter the sales records. When forecasting models are fed only with realized sales, this lost demand is systematically under-forecast. For this reason, demand planning uses a definition of demand that records unfulfilled demand wherever possible. This approach is called unconstrained demand.

Many factors affect customers’ demand and therefore a company’s sales. These factors should be considered within demand planning activities.

Demand planning is not a one-off calculation. It is a process that repeats at regular intervals. Forecasts are updated with new data in every planning cycle, results are compared with actuals, and the lessons learned are carried into the next cycle. Because sales, marketing, finance, production and procurement all provide input to the process, demand planning is also a collaborative process.

What Factors Affect Demand?

To prepare a reliable demand plan, input comes from many different factors: data, information and events compiled from internal and external channels.

Internal and External Factors

We can organize factors into two main groups: internal and external factors. The tables below summarize these factors.

Internal Factors
Historical demand and sales Pricing policy Delivery performance
Quality and rumors about quality Logistics and production costs New product launches, delists
Products' life cycle phases Sales and marketing activities Promotions and campaigns
Advertising, ATL/BTL, etc. Social media activities Management directives, quotas, targets
External Factors
Existing, new and potential customers Seasons, important days/weeks Weather conditions, climate change, natural disasters
Price changes and trends Market contractions or expansions Economic contractions or growth
Competitors' products and prices Competitors' new products Competitors' delivery performance
Competitors' sales and marketing activities Customer satisfaction and preferences Customers' investments and actions
Activities of customers' competitors Activities of customers' customers Major projects
Tenders and bulk purchases Raw material and commodity prices Overall logistics and production costs
Political and cultural factors Exchange rates Standards
Regulatory changes VAT, special consumption taxes, price indices, etc. Pandemics

Companies can add many more factors to the list above. The large number of potential drivers makes accurate forecasting more challenging.

Within demand planning, it is critical to identify, group and systematically monitor the elements that influence demand.

Classifying and Selecting Factors

Factors can also be classified by how far the organization can influence them. Price, promotions, advertising and new product launches are under the organization's control and are used to shape demand. Customer projects, tenders and competitor moves are partly predictable but outside the organization's control. Weather, exchange rates and epidemics are both outside its control and hard to know in advance. This distinction helps decide which factors enter the forecasting model as inputs and which are handled in scenario studies.

For a factor to contribute to the forecast, three conditions are sought: its effect on demand must be measurable, its future value must be known or reasonably predictable at the time the forecast is produced, and its data must be regularly accessible. The planned promotion calendar is a known input for future periods. Future weather can only be treated probabilistically. When a price change and a promotion take place in the same period, the effects of the factors must be separated from each other. For this purpose an events history is kept on record. For the calculation of promotion effects, see How to Forecast Promotion Impact.

Do Demand Forecasts Affect Competitiveness?

Demand Forecasts and Demand Management

The forecasts produced within demand planning activities are called Demand Forecasts. The activity of producing forecasts is called Forecasting.

Forecasting determines expected demand volumes and price estimates for different segments or groupings for future periods.

Organizations that do not generate forecasts are not planning their future effectively.

Forecasts are projections about how the future will unfold. Medium- and long-term corporate planning should be based on forecasts produced by demand planning.

Price changes or promotional activities intended to influence demand will also alter forecasts. When these actions are included in the planning process, the whole activity is often referred to as Demand Management.

According to Philip Kotler, "Demand management is influencing the volume, timing and mix of demand in line with corporate priorities and objectives."

Whether we call it Demand Planning or Demand Management, generating reliable forecasts through systematic processes is critical to corporate success.

Effect of Forecast Accuracy on Competitiveness

Companies that can plan demand better than their competitors will be able to plan and configure all supply chain operations - procurement, production, warehousing and logistics - more effectively. They can offer more reliable lead times and delivery windows to customers and gain competitive advantages through improved costs and prices. Those advantages lead to steady growth in market share and customer base.

In fact, a well-designed demand planning system is among the prerequisites for sustainable profitability.

The Cost of Forecast Error

The cost of forecast error arises in two directions. When the forecast is above actual demand, excess inventory, tied-up working capital, storage costs and obsolescence risk arise. When it is below, unfulfilled orders, lost sales, emergency production and emergency shipment costs arise. Because safety stock levels, production schedules and supplier orders are set according to the uncertainty in the forecast, an improvement in forecast accuracy is directly reflected in inventory and service level indicators. To see how forecast error is measured, see How to Measure Forecast Error: Bias and Percentage Error Metrics. For its relationship with safety stock, see 5 Common Mistakes in Safety Stock Calculation.

We observe that many companies with production or procurement problems actually fail at demand planning.

How Are Demand Forecasts Generated?

While designing demand planning processes, a systematic approach should be followed. A typical sequence can look like the steps below.

1. Define the demand planning process.

Clarify the importance of demand planning for the organization, its role in corporate planning, inputs, outputs, objectives and KPIs. Determine the departments and roles involved.

The process owner, the approval authority for forecast versions and the meeting calendar are also determined at this stage. Typically the demand planning function produces the statistical forecast. Sales and marketing propose adjustments using customer, campaign and price information. Finance and procurement evaluate the results in terms of budget and capacity. If it is not defined in advance who may intervene in the forecast at which stage and how the intervention is recorded, forecasts begin to be used with different figures in different departments.

2. Identify required data sources and systems. Collect the data.

Specify the drivers to be tracked, their sources and the ERP or other information systems that will be used.

When collecting data, level of detail and consistency take priority. It is checked which of sales, shipment and order data will be used as the basis for demand, how returns and cancellations will be handled, and whether product master data (product hierarchy, equivalent products, code changes) is up to date. Storing past forecast versions should also be designed at this stage. In systems that keep only the current forecast, past forecast error cannot be calculated afterwards.

3. Analyze data to choose forecasting preferences and methods.

Decide the level of aggregation for forecasts (products, subgroups, families, categories, brands) and whether to segment by location, region or customer. Determine forecasting horizons, update frequency and other process rules. Historical data (sales, shipments, orders, etc.) should be cleaned and enriched with an events history when necessary. Evaluate numerical forecasting methods (time series analysis, regression, machine learning, etc.). Consider judgmental adjustments where appropriate. Forecasting methods are detailed later in this article.

The method is not chosen with a single approach for all products. For high-volume products with stable demand, time series methods give reliable results. For items with intermittent demand, special methods such as Croston-type methods are used. For new products with no history, approaches based on the history of similar products or on managerial judgment are used. ABC-XYZ analysis, which classifies products by value and demand variability, helps determine how much effort each item deserves.

4. Evaluate and analyze forecasting performance.

Define performance indicators for forecast accuracy and monitor forecasting performance. Carefully choose error metrics (Mean Absolute Percentage Error, Weighted Mean Absolute Percentage Error, Bias, etc.). Evaluate overall process performance, gather feedback and propose improvements.

Accuracy is not measured with a single indicator. Indicators that show magnitude are monitored together with Bias, which shows direction. In addition, Forecast Value Added (FVA) is used to assess whether each intervention in the statistical forecast improves accuracy. For calculations and interpretation, see How to Measure Forecast Error: Bias and Percentage Error Metrics.

5. Integrate forecasts into corporate planning processes.

Include forecasts in operational, tactical and strategic planning processes. At these stages forecasts may be adjusted. later sections explain how forecasts are used in practice.

It is important to agree on a single consensus forecast figure in planning meetings. If sales, finance and operations work with different forecasts, inventory and capacity decisions lead to conflicting results. Once the agreed forecast is transferred to the planning systems, it is stored so that it can be compared with actuals in the next cycle.

What Methods Are Used In Forecasting?

Analytical and Judgmental Methods

Forecasting methods fall into two main categories:

  1. Analytical methods that produce numerical forecasts based on data.
  2. Judgmental methods that rely mostly on managers', employees' or experts' knowledge and experience.

Judgmental methods include sales force opinions, executive opinions, expert panels (the Delphi method), market research and drawing on the history of similar products. These methods are particularly valuable for new product launches, large customer projects and situations where historical data is insufficient. However, systematic tendencies such as optimism, pressure to comply with targets and giving too much weight to the most recent event can distort judgmental forecasts. For this reason, the contribution of interventions to accuracy is monitored with Forecast Value Added (FVA).

Statistical and Machine Learning Methods

Examples to analytical methods are listed in the table below.

Averaging Methods
  • Simple averages
  • Simple moving averages
  • Weighted moving averages
  • Moving medians
Time Series Methods
  • Exponential smoothing (ES)
  • Holt’s method (double ES)
  • Adaptive exponential smoothing
  • Holt-Winters (additive & multiplicative triple ES)
  • Seasonal indices
  • Decomposition analysis (STL, etc.)
  • ARIMA
Special Methods
  • Bass diffusion model
  • Croston method
  • Prophet (by Meta/Facebook)
Machine Learning Methods
  • Linear regression
  • Support vector regression (SVR)
  • Decision trees & random forests
  • Artificial neural networks (ANN)
  • Deep learning (LSTM, RNN, etc.)
  • Gradient boosting (XGBoost, LightGBM, etc.)

Method Selection and Comparison

In method selection, more complex methods do not always give more accurate results. The amount of data, the demand pattern (level, trend, seasonality, intermittency), the forecast horizon and the number of products are evaluated together. The general principle is to start from a simple reference method and test whether a more complex method beats this reference by a meaningful margin in accuracy. The last period's value or the same period of the previous year can serve as the reference. Methods are compared by holding back the last part of the historical data (a test period) and testing the methods on it.

For time series methods, Time Series Analysis for Demand Forecasting provides detail. For forecasting promotional periods, see How to Forecast Promotion Impact.

Which Corporate Processes Need Demand Forecasts?

Demand planning processes should produce the forecasts required by an organization for long-, medium- and short-term planning.

Hierarchical Demand Planning Levels

Long-Term Forecasts

Long-term forecasts are typically strategic forecasts for 2-10 years and are used in strategic business and investment planning.

Medium-Term Forecasts, Budgeting and S&OP

Medium-term forecasts cover 12-24 month horizons and are grouped forecasts for product or service families. They are used to determine procurement, production, warehousing and logistics needs and potential bottlenecks. These forecasts are prepared within Budget Planning or, Sales and Operations Planning (S&OP) processes.

Forecasts at different horizons must be consistent with one another. The forecast at product family level must not conflict with the sum of the forecasts at item level. This consistency is achieved with the top-down approach, which produces a forecast at a high level and distributes it to lower levels, or with the bottom-up approach, which produces forecasts at item level and sums them. Forecasts at a high level fluctuate less and are therefore usually more accurate, while item-level behavior is captured better at the lower level. In practice the two approaches are used together.

Demand Planning is one of the most critical stages of the S&OP (Sales & Operations Planning) process - also known as Integrated Business Planning (IBP).

S&OP is an integrated planning process that balances supply and demand over the medium and long term while focusing on competitiveness and profitability.

By improving cross-functional collaboration, S&OP helps organizations create plans that help them achieve corporate goals and respond faster to market dynamics.

  To read our article on S&OP fundamentals, click here.

Short-Term Forecasts and S&OE

Short-term forecasts are detailed forecasts for up to 2-3 months horizon and are used for operational procurement, production, storage and logistics planning. We refer to operational execution tightly linked to S&OP as Sales and Operations Execution (S&OE).

In the short term, forecasts can be updated frequently with current signals such as point-of-sale and order data. This approach is called demand sensing. The contribution of adjustments made with current signals to accuracy should also be measured by comparing them with actuals.

  To learn more about production planning, click here.

  To learn more about material planning, click here.

  To learn more about warehouse & logistics planning, click here.

Collaboration with Supply Chain Partners (CPFR)

Beyond internal S&OP and S&OE integration, Collaborative Planning, Forecasting and Replenishment (CPFR) among supply chain partners - sharing forecasts and plans - can drive excellence.

Why Must Demand Planning Be Tailored to the Organization?

There are standard practices for demand planning, especially for statistical forecasting. However, sales & marketing strategies, product/brand management, production strategies, supply chain operations and relationships, the nature of customer demand, interpretation of historical data, and assumptions about future actions mean demand planning must be tailored to each organization.

These factors and priorities require demand planning systems to be designed like bespoke solutions for each company.

For designing the demand planning process around the organization, selecting forecasting methods and performance indicators, and embedding the process in daily work, Demand Planning Consulting is available. For integrating demand planning with supply and financial plans, S&OP/IBP Consulting is available.

Tailored Demand Planning Processes image: pexels.com

References

What Demand Planning Services Does Optimum Planning Offer?

Optimum Planning provides practical training and consulting services to configure demand planning and forecasting processes and related tools tailored to your organization.

Starting with in-house workshops is recommended to embed the demand planning mindset and align stakeholders.

Our demand planning services and related offerings include the topics below.

Trusted By

Testimonials

Petrol Ofisi logo
“We had a productive training for two days. As a company, we have seized opportunities to see our shortcomings and make improvements. Thank you very much.”
Yiğit Doruk Turguter, Senior Engineer in Production and Demand Planning, Petrol Ofisi (Demand Planning, 2024)
Sarsilmaz Firearms logo
“It was a training that was useful for our sales and planning group. Thank you.”
Murat Işık, Planning & ERP Manager, Sarsilmaz Firearms (Demand Planning, 2023)
“It is a training that every company operating in a geography like Turkey, where uncertainty is increasing day by day, should receive. I think that this training will provide a great advantage in our business processes, as it is very critical to make the right decisions in an aggressive competitive environment and to make the right predictions in parallel.”
Hale Topaloğlu, Senior Product Manager, Global Pharmaceutical Company (Demand Planning, 2023)

To view all testimonials, click here.

To arrange an in-house training for your organization or to join a public session, request our proposal.

Request Proposal  
Back to top