Data analysis
Our project managers master all the current multivariate methods as factor, cluster, discriminant and regression analyses. We are also familiar with the various methods of conjoint-analysis as well as with price research.
Decisive for the choice of the method is the project goal – analysis methods are not an end in themselves. Analysis methods are about answering the questions raised and obtaining practical, concrete results. We decide anew for every single project whether a frequency analysis is sufficient or moreover multivariate analyses are deployed.
To deliver only a volume of tables of processed final results is not the goal of our work. We offer more to you – results depicted and structured in a tangible and clearly-exemplified graphical form. Furthermore, complex circumstances are focused to the customer point of view.
In that connection, we describe not only what happens in the market but also why. We place a special value on performance-oriented and clear recommendations as well as concrete actions for strategy, communication, service, marketing and sales.
Übersicht der von COBUS verwendeten multivariaten Methoden (Auszug)
Methode | Funktion, Nutzen |
|---|---|
Faktorenanalyse, Multidimen-sionale Skalierung (MDS) und Korrespondenzanalyse | Strukturieren und Zusammenfassen mehrerer Skalen auf wenige Dimensionen (z.B. „Mental Maps“), Positionierungs-analysen |
(Logistische) Multiple Regression | Wirkung mehrerer Faktoren („Prädiktoren“) auf ein interessierendes Kriterium (z.B. Kundenzufriedenheit) |
Clusteranalyse, Latente Klassen-analyse, Diskriminanzanalyse | Segmentierungen, Bestimmungen von Typologien |
CHAID Analyse | Entdecken komplexerer Zusammenhänge zur Erklärung eines Merkmals („Entscheidungsbäume“) |
Conjoint-Analyse | Produkt- und Preisanalyse („Dekomposition von Produkten“) |
Struktur-Gleichungsmodelle, Pfadanalyse, Kausalanalyse (LISREL, PLS) | Überprüfen komplexerer Ursache-Wirkungsgefüge, ggf. unter Berücksichtigung latenter (nicht direkt beobachtbarer) Variablen und gleichzeitig Tests der Güte von Modellen |
Neuronale Netzwerke | Dynamisches Erlernen von Assoziationsregeln, z.B. zur Abbildung von Lernprozessen von Verbrauchern |
Conjoint-Analysen
In the last years conjoint analyses have become an increasingly important component of market research. Those allow the individual assessment demand preferences and thus the implementation of constructive segmentation in the field of competition and the extrapolation of propositions on the likelihood of success for alternative marketing strategies.
Conjoint analyses are particularly suitable for
- Verifying the maturity in the market of already established products / brands
- Verifying to what extent are new products ready fort he market
- Verifying the outcomes of changes in the product
- Verifying the acceptance towards new product variants
- Estimating the market share for new products versus those of the competition
- Price analysis – Assessment of the price which maximises profit
We generally offer guidance about targets, fields of application, methods and about the conjoint analysis process. We establish the most suitable procedures, the appropriate product features and the design (Choice Based Conjoint CBC or Adaptive Conjoint Analysis ACA). We carry out the research among fixed target groups. Subsequent to the fieldwork phase, the supplied data is processed by means of the conjoint analysis and the individual user values are analysed.