Identify and measure key drivers of customer behavior impacting the user satisfaction for a Digital Electronics device.
Our standard solution has been Max-Diff, we used Key Driver on your recommendation. And it was bang on - we delivered exactly what our client was after.
- Project Manager, Research Excellence, Market Research Consulting, USA.
A mid sized market research consulting company is looking to go beyond the standard pieces of analyses to distinguish what really are the key factors causing satisfaction in usage amongst the consumers of a digital electronics product.
The product is a uniquely positioned item that has been attacked by similar products over the past few years, but has retained its charm amongst the customers due to brand recall and looks. Despite some seasonal drop in the sales, this has largely remained the company’s star product with capturing top position in the competitive landscape.
The company is now looking to understand the ‘sharing’ aspect of the product for other products to mimic and reflect. What leads to the customers being so satisfied and recommending it to potential customers and acquaintances remains a key challenge to understand.
Structuring the Business Problem: Key factors (or drivers) of satisfaction are to be understood. To structure the study, every variable and the response distribution is studied to look at the landscape of the data we have available with us. Open ended questions are given lower priority than questions with fixed responses. One of the questions is found to fulfill the criteria on which we could center our analysis. It had a 10-point response sheet and only a single answer is to be allowed hence we have overall satisfaction scores available with us on a scale of 1 to 10 :
This question is taken as the dependent variable while responses to two questions are chosen as the independent variable due to the exhaustive list of factors available with us:
In all, there are 42 independent variables available with us. Almost all the respondents have filled the survey quite well and only eleven respondents are found to fill the survey with missing values. Analysis Method: Because of the exhaustiveness of the factor variables, we identify the possibility of highly correlating variables within the regression models. We hence use various statistical techniques to cluster similar attributes out of which only one is to be chosen for the final model. Multiple iterations of models are conducted and results discussed with the research team to fix the final set of attributes that were aligning with the business rules. Model is developed and drivers are selected for presentation and discussion with the research team. (more details in Appendix 1)
Using the analysis, we are able to identify key drivers of satisfaction amongst the consumers using the following rules:
1. High Drivers - which have highest ‘Relative Importance’ & Significant p- values. Rest of the Drivers sorted on their correlation with the ‘Overall satisfaction Score (Q25)’.
2. Medium Drivers – Correlation with ‘Overall Satisfaction Score (Q25)’ is greater than 0.3 & ‘Relative Importance’ less than 5%.
3. Low Drivers - Correlation with ‘Overall Satisfaction Score (Q25)’ less than 0.3 & ‘Relative Importance’ less than 2%.