Analysis methods
The questions are the foundation of market research. The true art however lies in transforming the answers given by the sample into powerful decision aids. The individual tasks determine the appropriate method, i.e. when multivariate processes should be used, when structure revealing analyses should be carried out and when a variance analysis comparing different levels would be ideal.
All methods and processes are familiar to INNOFACT AG. Our team brings competence to even the most complex analyses and our actions focus on the optimal solution.
Of course INNOFACT AG commands the specialist knowledge that can be gained at university and applied professionally. Furthermore however, we have developed known methods further, on which sound answers and economic success can be based.
You can obviously expect descriptive evaluation and standard statistical methods from us. By tradition however we feel at home with procedures and applications that go that one step further, for example:
 | Conjoint: Our experience of many years enables us to discuss projects cleanly and quickly and come to a confident decision about the appropriate study design. It is not necessary for us to make any preliminary tests or trials; instead we focus immediately on the point. We have a comprehensive software pool for both full profile and adaptive conjoint analysis. |
| Structure revealing procedures and typology serve to illustrate the relationships in the data which are not immediately recognisable. Often we seek to segment test persons i.e. to find interviewees similar to each other and fundamentally different from other groups in society. |
Imagine that you are the marketing manager of a chain of electronic retailers and with us you conduct a market research study analysing customer satisfaction. In each case we aim to organise customer structure. It is probable that after close inspection three or four fundamentally different types of customer will emerge, for example the technology avoiders who see the purchase of electronics only as a necessary chore and perhaps want to buy nothing more than a replacement part for a broken fridge. On the other hand we could probably identify a group of customers falling into the category "technology lovers". The groups behave in completely different ways interesting to marketing managers. The groups visit such shops with differing frequencies and are easier or more difficult to reach through advertising (newspaper/internet) etc. We take a close look at each case and are therefore able to deliver much better information than the market research that you probably already know.
| Hypothesis Tests, in contrast to structure revealing processes, are ideal when precise information is required, for example if you wanted to know whether a new label on a beer bottle effected a higher buying incentive as the old one. The number of hypothesis tests and the variations thereof are limitless - from methodical, relatively simple t-tests and variance analyses, through to non-parametric tests and lengthy multivariate covariance analysis. As statisticians with experience spanning many years we not only command these processes, but we also know how to administer the appropriate "doses". You can therefore be sure that we do not overuse the methods, but that we always apply them where appropriate and useful. |
 | Back |  | Top |  | Contact |
