modelQED Overview...
What Is Marketing Mix Modelling?
modelQED Key Features...
Which Metrics Can We Analyse Using modelQED?
How modelQED Helps You Calculate Marketing ROI
Automating modelQED For Quicker Marketing Analysis?
modelQED Frequently Asked Questions...

What Are The Key Features Of modelQED^®?

modelQED is structured into three key areas - each containing a range of features designed to improve the efficient construction of marketing mix models:

• Basic Analysis - review data, produce key graphs and tables.


• Modelling - create marketing mix models (econometric models) visually or via automated generation engine.


• Model Management - review all models which have been saved within your project.

Basic Analysis

The data analysis section of modelQED is the area where data can be imported, analysed, manipulated and charted. modelQED is designed to make importing data a quick and painless task - it imports directly from flat files that can be created in almost any package or database to ensure maximum interoperability across platforms. Once data has been imported into modelQED, users need to identify each series in order to enable modelQED to model the data accurately and efficiently.

modelQED also offers a large number of data charting and analysis screens that allow users to analyse the data in their data set. Charting options are extensive and flexible whilst analysis screens cover some basic statistics such as correlations, autocorrelations, partial autocorrelations and cross correlations.

Modelling

The modelling section of the modelQED application lets users create marketing mix models using either a "manual" or an "automated" methodology. The "manual modelling" option allows users to create a model in a "point and click" environment giving users an unprecedented modelling experience. Users have full control over variable selection, transformation (including adstock and curve types) and lag length selections within a model and the results are immediately translated into meaningful reports which are "easy to interpret" - not just in a statistical context but in a business context.

The "automated" modelling procedure uses advanced user-defined modelling algorithms to search a given data set for the "best" model or models which meet the analysts need. Users propose a set of data for examination by modelQED and given this set of data, seek for the combination of variables, adstock transformations (advertising half life), curve transformation and lags which best explain the dependent variable (i.e. the variable that the business is interested in understanding).

Whether a model was created using the manual or automated methodology, modelQED seeks to make model interpretation as easy as possible. Most common model diagnostics are pre-calculated saving hours of spreadsheet analysis - this makes it very easy and quick to get to the analytical answers that are needed.

The range of reports created for every marketing mix model produced feature both statistical outputs and business reports. Of these, the most important reports are those which calculate the various definitions of "return on investment" generated by each driver within a model. modelQED goes beyond almost all other packages in this respect by reporting marginal returns in addition to the more basic average levels of return on investment.

Model Management

The model management section of modelQED allows users to compare and review saved marketing mix models. Project and Variable Filters enable users to compare results across a wide range of criteria ensuring that the best available models are those which are reported on. Finally users can re-edit saved models - even if they have deleted historical data from their main data set.