Forecasting

Forecasting is a methodology in operational business planning and control, whereby the resulting forecasts allow an assessment of whether good times can be expected or crises need to be managed. The method also serves for damage control.
Forecasting

What is Forecasting?

Forecasting is an instrument for controlling and planning in business management. The term can be translated as prediction or forecast. For example, companies use forecasts to plan monthly budgets for costs, revenue, profits, and liquidity. Factors such as liquidity are essential for businesses, and the relevant developments must be monitored regularly.

Otherwise, companies run the risk of being unable to plan personnel assignments, miscalculating marketing costs and other expenses, or not allocating budgets efficiently. With forecasting, analyses are developed and trends observed. Without this, a company may suffer as existing resources may not be fully and profitably utilised.

An analysis or forecast regarding developments is an important instrument in controlling. Using a time series forecast, a key figure can be estimated as accurately as possible for the future.

What are time series?

The term “time series” originates from mathematics and describes the temporal course of observations and measured variables. Time series contain important statistical indicators that are arranged in time. Such a key figure can be the number of graduates from a specific education program, for example. When looking at these numbers over several years, this is referred to as a time series.

The prediction of future values of a measured variable based on time series is called time series forecasting or time series analysis. In English, this is referred to as Time Series Forecasting.

Frequently asked questions and answers  

Forecasting is based on analyses of historical data and key figures through an algorithm. This application of artificial intelligence recognises future developments based on so-called time series, thus enabling companies to have a certain level of planning security. To create a concrete forecast and make predictions, various steps must be taken.

Initially, key figures must be made available in the form of time series (data is observed over an extended period). Then the developments are described based on intelligent software. Back testing is also used to ensure that the model is meaningful for future developments and planning. In this way, future challenges may be overcome effectively.

Final forecasting models can generate new predictions at any time. The software solution is a long-term investment in consulting and managing a company.

With meaningful assessments, secure recommendations for the future can be made. To do this, data is generated and collected in forecasting. The resulting results can be guiding factors for companies. For example, deviations from existing plans, such as in production or transport, can be detected early. This saves costs and avoids bad investments.

Various methods are relevant for creating forecasts about the future. However, two main groups are particularly important: the quantitative and the qualitative method.

Various methods are relevant for making forecasts about the future. However, two superordinate groups are particularly important: the quantitative and the qualitative approach.

Quantitative Method

In this type of forecast, historical data from a company is used as groundwork. Using a mathematical formula, for example, the demand for a service or product in the market can be determined. Past sales figures and demand trends are particularly relevant.

How much data a company includes in the forecast is at their discretion. However, the more data used, the more precise the forecast will be. The quantitative method is therefore less suitable for young companies with little data.

Qualitative Method

In contrast to the quantitative method, the qualitative method primarily involves subjective assessments of experts, rather than large datasets. The future demand in the market is estimated precisely based on specialised knowledge and experience. To do this, companies can hire experts both internally and externally.

In the qualitative method, there is also a special form of forecast. Demand forecasting includes predicting how a sales promotion will affect demand and consumer behaviour.

Sources

  • Organisation für wirtschaftliche Zusammenarbeit und Entwicklung (OECD, englisch: Organisation for Economic Co-operation and Development). Glossary of statistical terms. Forecasting.
    https://stats.oecd.org/glossary/detail.asp?ID=3652