What are demand forecast methods?
Demand forecast methods refer to the ways through which an organisation make estimations about future customer demand. However, this concept is a subset of demand forecasting which will be discussed later in detail.
Demand forecasting explains analysing or predicting upcoming demand over a defined period using either historical data or other related information.
Proper demand forecast methods provide valuable information to the business about the other markets and potential in the current market. This is carried out so that managers could make decisions about business growth strategies, pricing, and market potential.
Without demand forecasting technique, the business ends up making poor decisions about target markets and products where ill-informed decisions possess negative effects on customer satisfaction, profitability, inventory holding costs and supply chain management.
Importance of demand forecasting
There are various reasons to explain as to why demand forecast methods are important to enrol in the businesses which are as follows:
- Demand forecast methods allow the business to optimize inventory, reduce holding costs and increase inventory turnover rates more effectively.
- Sales forecasts definition aids with budgeting, planning and goal setting. Once the organisation gets an estimate of future sales, an informed procurement strategy can be developed to match the customer demands.
- Through demand forecast, a company can find kinks in the pipeline before time for robust performance.
- Facilitates insight into future cash flow. This explains that businesses can prepare a budget more freely to pay other operational costs and suppliers and invest in the growth of the company.
- Anticipating demand leads to knowing when there is a need for an increase of staff and other resources for the smooth running of business during peak periods.
Methods to Forecast demand
Different organisations use different methods for demand forecasting technique depending upon the nature, requirements, budget and size of the business for a future period. Broadly classifying there are two categories to define demand forecast methods which are discussed further in detail.
- Qualitative method
- Survey method
- Opinion Polls
- Quantitative method
- Smoothing technique
- Econometric method
- Time series analysis
- Barometric method
The qualitative methods to forecast demand focus on the buying behaviour of the customers to collect the data either through conducting surveys or gathering from experts the methods to forecast demand. This technique is usually considered to make a short term demand forecasting technique.
This technique is generally facilitated in situations when historical data is not available such as while the introduction of a new product or service. Qualitative methods are based on judgement, experience, conjecture, intuition, etc.
The most commonly used technique in the short run is the demand forecasting survey methods. They reply to future purchase plans and intentions of the consumers to anticipate demand. Therefore the surveys or search are conducted with customers to find out their existing products demand and future relation accordingly. Furthermore, are two kinds of surveys are explained:
- Complete enumeration: Also termed as census method under demand forecasting survey methods. Here, most of the potential customers of the product are contacted and analysed about their buying plans or behaviour. Based on them the predictions are made. The aggregate is however known by totalling the probable demands of all individual consumers in the target market.
- Sample survey: In this type of demand forecasting survey methods, few potential ones(called sample) are selected from the whole segment or market and surveyed. The average demand is guessed or analysed based on the research gathered from these samples.
Opinion poll includes taking the suggestion of who possess knowledge of market trends like marketing experts, sales representatives and consultants. The commonly used opinion polls are as follows:
- Market studies and experiments: Also referred to as market experiment method. In this technique, the company initially sets certain aspects for a market like income levels, occupational distribution, population, cultural or social background, and customer’s tastes and preferences. Among them, one is selected, and its effect on demand is predicted while keeping all other factors constant.
- Expert opinion: For this technique, the sales representatives of the different company connect with customers buying behaviour, responses and reactions to changes, opinions for new products introduced etc. Then the prediction of experts are analysed and demand forecast methods decisions are made.
- Delphi method: Here, market experts are facilitated with assumptions and estimates of forecast searched by other experts in the industry. Experts can revise or reconsider their work or decisions based on information provided by others.
As the name suggests, the demand forecasting quantitative methods use statistical tools as demand forecast methods. Here, the demand is forecasted based on the historical data provided. Vice-versa of qualitative method, they are used for long term forecasts.
Statistical tools are reliable and cost-effective since the contrast of subjectivity is minimum as possible. Therefore, let’s discuss the different kinds of demand forecasting quantitative methods:
Time series analysis
Time series analysis also termed as a trend projection demand forecasting quantitative methods that turn out as the most popular demand forecast methods used by companies in the long run. The concept of time series explains the sequential order of values of a variable (called trend) for equal time intervals.
Predicting trends, a company can forecast the demand for its existing product and services. There are four components that a company should take into consideration of time series analysis while demand forecast methods which are as below:
- Cyclical component: This accounts for a regular pattern of sequences in time series analysis of values below and above the trend line lasting more than one year.
- Trend component: It counts the gradual shift in the time series analysis to a relatively lower or higher value over some time.
- Irregular component: This kind of demand forecast methods is for short term, non-recurring and unanticipated factors affecting the values of time series.
- Seasonal component: Seasonal time series analysis determines regular patterns of variability within a certain period like a year.
In situations where time series analysis does not work or lacks trends, smoothing techniques are used. They are facilitated to eliminate random variation from the historical demand. Guides with identifying demand levels and demand patterns used to predict future demand. Common demand forecast methods of smoothing technique are weighted moving average method or simple moving average method.
- Simple moving average method: It is determined by calculating the mean of average prices over a period and plotting them on the graph which acts as a scale. For example, the six-day simple moving average counts as the sum of all six days divided by six.
- Weighted moving average method: It accounts for the use of a predefined number of periods to find out the average, all of which have the same significance. For example, in a four-month moving weighted average, every month represents 25% of the average.
The barometric method analyses future trends based on current developments in the company. This method is also termed as leading indicators approach for demand forecast methods.
A variety of economists use the barometric method to predict trends in the business. The key idea followed in barometric demand forecast methods is to prepare an index of economic indicators and predict future trends based on movements depicted in the index.
The barometric method uses the following indicators:
- Coincident indicators: These indicators move correspondingly with the current event. For example, the rate of employment, number of employees in the non-agricultural sector, per capita income etc work as indicators of a nation’s economy for the current state.
- Leading indicators: The event that happened already is considered to forecast the future event, where the past event act as a leading indicator. For example, the working women related data act as a leading indicator for the demand of women working in hostels.
- Lagging indicators: Lagging indicators involve events that follow a change. These indicators are however difficult to interpret how the economy can shape in future. But useful in analysing future economic events. For example, unemployment levels, inflation etc can be quoted as indicators of the performance of a country’s economy.
The econometric method uses a demand forecasting statistical methods tool along with a combination of economic theories to assess the number of variables (like income level of consumers, price change, changes in economic policies etc) for demand forecast. The predictions made using the econometric method are more reliable as compared to other demand forecast methods.
The model for econometric demand forecasting statistical methods can be a single equation of regression analysis or a system of multiple equations. The concept of regression can be discussed as follows:
- The regression analysis measures the relationship between two variables. Using these demand forecast methods a relation is established among independent (advertisements, an income of consumer, and price of related goods etc) and dependent variable ( quantity demanded).
- In other words, it is a demand forecasting statistical methods tool to find the unknown value of a variable when the independent variable is known. After creating the relationship, the equation is derived assuming their relationship is linear.
- Y= a + bX, where “Y” is the dependent variable for which the demand forecast methods are used. The “b” is the slope of the regression curve while X is the independent variable and “a” being Y-intercept.
Examples of Demand forecast methods
Once the company has the basis of the above demand forecast methods, they can define what is their requirement and use that technique accordingly. Following are few examples of the benefits for demand forecasting:
The company Zara follows a just-in-time production approach where they manufacture, design, distribute or sell within two weeks. The company keeps large production in-house, to increase the flexibility of the production cycle, supply chain and control over the manufacturing process over competitors.
The sales and customer feedbacks are facilitated to designers to make adjustments as quickly as possible and line up with customer demand. They also have extra labour capacity to meet the demand in shifts stabilizing the company’s lean inventory management approach.
In the year 2001, the company Nike installed demand-planning software without testing, resulting in an overstock of shoes (low-selling) and an understock of popular Air Jordans. As a result, it cost $100 million worth of sales to the company.
This case is an example of lost out when they tried to implement a new system too quickly. However, the demand forecast methods are essential for predicting sales or managing inventory. But the introduction of any new system requires rigorous testing before being rolled out.
Ikea inventory management strategy facilitates an inventory system of proprietary. That works as a guide for logistics managers with point-of-sale and warehouse management system data. The company outlines multiple inventories that comes into the store through distribution centres and direct shipping.
This information leads the managers to predict sales for a couple of days and order products to meet the future demand. However, if the sales do not match with the project turnover of that day, he counts manually the products in stock. As a result, an excellent example of demand forecast methods with a two-step verification of manual process acting as safety for accuracy.
Walmart’s supply chain is understandably complex. But as their logistics are known for being technologically advanced and precise. In 2013 there was a serious problem of in-store out-of-stock.
The company’s understock on shelves was determined by mismanaged inventory- explaining stock was present in a warehouse. But, not enough staff was available to move them on shelves. In this situation, cost-cutting reflected a negative customer experience, that could have been avoided by proper demand forecast methods.
The demand forecast methods, therefore, explains how to find in advance the expected requirement of the current product. There are various techniques available depending upon the choice of the company to use. Which is best suitable according to the size and nature of the organisation. Also, there are few real-world examples to explain the dos and don’ts while using them.