Prozentor  >  Forecasting  >  Market Research

Market Research Forecasting

Accurate market analysis can contribute substantially to long-term success; standard analysis methods do not always provide the right answers.

Utilising a system of proprietary methodological components we are able to develop custom analysis systems for nearly any research application. Market research data are very well-suited for time series analysis, and for forecasting in particular. The typical characteristics of market data such as trend, season and economic cyclicality are factored in.

Why conduct such forecasting?

This type of forecasting is useful for obtaining answers to marketing-related questions and helps users make better-informed media decisions.

Common research objectives are:

• Optimisation of advertising spending in the upcoming quarter
• Determining what phase the advertising market is in
• Gauging what changes will there be in advertising spending in my industry over the months ahead

What is special about this type of forecasting?

This type of forecasting takes account of both seasonal and exogenous influences. A dates database ensures that major events like the Olympic Games and calendar effects like variable Easter holidays are factored into forecasting.

For example

Statistical time series analysis allows objective model-based estimates, providing for high forecast quality.

Who uses this kind of forecasting?

Advertising managers and media agencies utilise market forecasts to optimise their own marketing and advertising behaviour. Industry, industry segment, product group and competitors can be analysed. Our market forecasting can uncover optimisation potential regarding share of advertising, seasonal budgeting and composition of the media mix.

It also has useful applications for media providers and their marketers. Our forecasting makes it possible to optimally structure future advertising sales. The key concepts here are strategic optimisation of the product portfolio, advertising offer optimisation, sales management and validation of budgeting and sales projections.

Procedures and methods

1. For every market research time series, SARIMAX models (univariate and multiple time series analysis) are used to generate a multi-step forecast.
2. Then the hierarchical structure is mapped using linear sub-conditions in a specially developed procedure.
3. This yields consistent forecasts for the entire scope of market research. Forecasts for the entire range of industries for example add up to constitute an overall market forecast.