Forecasting the Future: Types, Techniques and Methods of Demand Forecasting
Planning is at the forefront of success for any organisation, big or small, commercial or otherwise. What about the supply chain that involves a huge flow of goods, revenues and a company’s properties? Turns out it is critical here. When companies strategically plan for the future by anticipating possible trends in sales, customer purchases, and required inventory, we call it demand forecasting in the supply chain. It is also known as sales forecasting.
In fact, it involves more factors than these three. From resource allocation, budget planning, and supplier onboarding to setting a company’s whole trajectory, sales forecasting plays a major role. How effectively organisations can forecast depends on the methods used and market awareness. So, in this article, while touching on the importance of demand forecasting in supply chain, we guide you through the popular types, techniques and methods of Demand Forecasting
Why Demand Forecasting
- Inventory Management: Ensures optimal stock levels, preventing overstock or shortages.
- Resource Allocation: Helps allocate resources efficiently based on anticipated demand.
- Cost Optimisation: Minimises inventory storage costs, reduces wastage, and enhances operational efficiency through planning.
- Customer Satisfaction: Enables timely order fulfilment and deliveries for customers.
- Strategic Planning: Demand forecast enables long-term planning and decision-making, which are crucial for sustainable supply chain operations.
- Supplier Relationships: Enhances timely collaboration with suppliers by providing accurate demand information.
- Competitive Advantage: Helps adapt to market changes and fluctuating customer demands and stay competitive in dynamic environments.
Popular Methods of Demand Forecasting
There are many demand forecasting techniques in use. They can be broadly divided into qualitative and quantitative methods. Quantitative methods use mathematical techniques such as statistics, probability, etc., in the process of demand forecasting. While many companies rely on the combination of two techniques, the choice depends on the scale of business, type of product, forecast period, and data availability.
Market Research
The most common method of assessing future demand is by directly asking the customers. Of course, indirect methods such as online customer surveys, social media engagement, forums, etc., can reveal new and often surprising insights into what customers want that are not figured in the sales data. Demand forecasting through market research informs about new marketing efforts, customer expectations and needs, and opportunities.
Sample surveys, complete enumeration surveys, and end-user surveys are some of the common tools used for market research.
Trend Projection
It is a quantitative method of demand forecasting that uses statistical analysis or regression analysis. It is actually a simple and cost-effective method based on past sales data. Predictions are made by extrapolating the data into the future. The basic steps are:
- Collect historical demand data over a suitable period.
- Plot the data on a graph to visualise the trend.
- Use analytical techniques like regression analysis or statistics to quantify the trend line that best fits the historical data points.
- Extrapolate or extend this trend line into the future to project future demand levels.
Different types of trend models are used, such as linear, exponential, quadratic, etc., based on the pattern observed in the data. Trend projection is useful when demand exhibits a clear upward or downward pattern over time.
However, it may not account for factors that could cause deviations from the past trend in the future. For example, when a product goes viral, sales may go up. But trend projection can not account for that.
Econometric
Econometric demand forecasting exploits the premise that causal factors drive demand. It uses economic, marketing, and other relevant data to draw a relationship between demand and the influencing factors. The complexity increases as the number of factors/variables increases. Forecasting in the premise of multiple variables presents nuanced mathematical equations.
Think about what happened during the COVID-19 pandemic. Many people started shopping online because of lockdowns, which became a trend gradually. Similarly, if the economy is doing well, people might have more money to spend on travel and vacations. But if people have more debts, it could be because they spent money on fixing up their homes.
Demand Sensing
An effective demand planning technique is demand sensing. It utilises machine learning to track real-time shifts in purchasing behaviour. AI and ML software contribute to establishing a data-driven supply chain.
It is advantageous as demand sensing enables quick responses to unforeseen demand changes, enhancing service levels and forecast accuracy. AI-enabled demand sensing tools, such as LEAFIO, Blue Yonder, and SAS, offer real-time visibility into short-term demand, further improving overall operational efficiency.
Use of Predictive Analytics
Using predictive supply chain analytics, you can forecast demand and figure out factors driving sales. The visibility comes from combining smart technology (like IoT) and machine learning, showing every step of the supply chain to predict future demand.
With good ML-powered tools, like Salesforce or Anaplan, you can aggregate old and new data from surveys, social media, customer opinions, and more to get a more solid demand forecast.
Delphi
It is similar to expert opinion, as the forecast takes the views of experts into consideration. However, it is only a small part. Delphi method is a process of demand forecasting where companies send surveys to industry experts across the world. After compiling the responses, they are again anonymously sent back to the experts for feedback. This process is continued for one or two rounds until a consensus is reached. The final results are used for the demand forecast.
Because surveys are sent anonymously, the Delphi method achieves honest expert opinions. Moreover, multiple compilations ensure that the result goes through refinement, building on each other’s knowledge and opinions.
The sales force composite is a way to predict how much of a product will be needed. It depends on the sales and marketing people as they know the market well. Sales and marketing teams in different areas share their predictions based on what they see in the market. All this information is put together to create an overall prediction for the whole company.
Sales Force Composite
However, companies often distinguish predictions based on factors like product price, marketing, customer wealth, and competitors. They might also be specific to a particular region and demographics.
Expert Opinion
While customer surveys and several other techniques can inform demand planning, expert opinions are invaluable. Their experience might help identify twists and common pitfalls in predictions and assumptions.
Companies that use expert opinion hire an outside expert to accomplish the task. The company and the experts sit together to brainstorm and come up with assumptions. Leadership then makes decisions based on the assumptions for the future.
Realted read: Understanding Demand Management
Other Demand Forecasting Techniques
A/B Experimentation
A/B experimentation or split testing is primarily used in marketing, website design, and product development to compare two versions (A and B) of a product, promotions, website, email subject lines, advertisement, etc., to see which one performs better based on specific metrics like clicks, conversions, sales, etc.
Based on a clear preference for one option, companies can enhance their understanding of consumer appeal, aiding in demand prediction. For instance, a study revealed that companies achieve higher sales by setting prices with odd-ending numbers!
A/B testing is useful in a limited way for testing promotions, pricing changes, etc., and measuring their impact on demand in a controlled environment. However, it is not considered a mainstream method of demand forecasting in supply chain management.
Barometrics
It is a very specific and short-term demand planning technique that uses certain indicators to predict trends. Mostly, these methods are used in inventory planning and supply chain management.
- Leading indicators are used for predicting future trends. For example, an increase in consumer complaints indicates a potential sales drop.
- Lagging indicators point at future trends based on past data. For example, companies can plan their inventory for the next month based on the sales spike in the previous month.
- There are coincident indicators based on current events. For example, tracking the inventory based on ongoing sales.
Types of Demand Forecasting
Active Demand Forecasting: It is a popular method for startups and growing businesses. It seeks to achieve aggressive growth by focusing on driving factors such as product development, competition, economic projections, market growth trends, etc.
Passive Demand Forecasting: This is the most general type of demand planning every established industry uses. It utilises past sales data to project future trends. Unlike active demand forecasting, the passive technique prioritises stability over growth.
Short-Term Demand Forecasting: When there are sudden surges in sales due to festivities or such factors, organisations must resort to short-term demand forecasts. So, companies can employ this technique while long-term forecast plans are already underway, to quickly adjust to the changing customer demands.
Long-Term Demand Forecasting: When companies project trends for the next one to four years, it is called long-term demand/sales forecasting. Typically, it sets the company’s trajectory for the set duration. Marketing campaigns, capital investments, and internal supply chain operations are some of the examples.
Internal Demand Forecasting: When forecasting demands, companies must also consider internal factors such as inventory, staff, production capacity, machine floor, etc. Past sales data is also helpful for this process. It is often part of strategic planning.
Macro Demand Forecasting: A macro-level sales forecasting model examines broader external factors like raw economic trends, material availability, the state of logistics, and others that impact various aspects of the supply chain. Considering these factors can help achieve accurate demand predictions.
Conclusion
Demand forecasting is vital in supply chains, aiding inventory management, restocking decisions, and capacity planning. For instance, warehouse fulfilment services rely on demand forecasts to determine storage space, product quantities, and staffing needs for efficient shipping and inventory replenishment. Accurate forecasting ensures optimal resource allocation and effective inventory management, without which organisations have a hard time serving customers.
FAQs
What are the two methods of demand forecasting?
There are two broad methods of demand/sales forecasting: Quantitative and Qualitative. While qualitative methods rely on expert opinions, market surveys, and judgmental inputs, quantitative methods rely on mathematical models drawn from past data.
What are the four steps in demand forecasting?
Identifying Markets.
Splitting the overall industry demand into segments.
Find out the cause of demand in each segment and what might change them.
Find out the risks of the forecast and finalise the most effective projections.
What is the oldest demand forecasting technique?
The oldest method of demand/sales forecasting is the “Consumers Survey Method”, which is also known as Opinion Surveys.
What is demand forecasting in economics vs supply chain?
In economics, sales forecasting predicts aggregate demand. It focuses on drivers like GDP and consumer behaviour. In the retail supply chain, it focuses on predicting demand within a company and for their products, to plan production, inventory, order fulfilment, and logistics well.
What is demand forecasting in managerial economics vs supply chain?
Demand forecasting in managerial economics analyses market demand facts like prices, income, and competition to optimise profit. In the supply chain, it predicts product/service demand for operational planning of procurement, production, and distribution to meet customer needs efficiently.
