Using decision management tools to improve research

Click here for a French version

Why do you use IBM SPSS solutions?

I use IBM SPSS software for my lectures, my research projects and for publishing my research. The software is really well suited to the university environment, mainly because it’s extremely user friendly and offers all the statistical functions we and our students need (standard deviation calculations, factorial analyses, etc.). Graduate students in Marketing use the IBM SPSS tools on which they were trained, although of course the software can never replace human interpretation.

When I use the software for my marketing research projects, I appreciate, as do many other researchers, that it’s a sophisticated software solution that’s made all the easier to use thanks to online help that can, for example, help you to understand something as straightforward as the significance of certain spreadsheets. This is a fundamental requirement for corporate marketing teams, because it’s not the job of marketers to understand how the software works, but rather to work effectively using simple, powerful tools that help them to make the right decisions!

For example, with IBM SPSS Modeler you can use data on survey results as the basis for preparing a senior management presentation very easily and very quickly … but, at the same time, those who want to go further can model the correlations between sales and explanatory sales-related variables (advertising spend, price, marketing promotions implemented during the period, etc.), which provide a clearer explanation of increasing or decreasing brand sales and more meaningful answers to the questions asked by senior management teams trying to understand and learn from experience.

Does research work impose the same requirements?

Yes, the problems and analytical requirements are similar: we need a system that’s simple to use but delivers comprehensive functionality. In the world of research, we use relatively complex models, especially when it comes to identifying commonalities in consumer behavior. We correlate sales with explanatory variables, but it’s often much more complicated than that (non-linear modeling, mix of quantitative and nominal variables, etc.). For example, we have just installed IBM SPSS Modeler to work on more complex models and predictive analysis as part of a major call center research project funded by Cap Digital.

Details of research projects are often published in specialized journals, and in time the results arrived at by academic marketing researchers often become the benchmarks and standards used by businesses; so it’s important for business managers to familiarize themselves with the work being done by researchers, especially in areas such as consumer behavior modeling and the ROI delivered by marketing campaigns. They often discover that academic researchers have already answered some of the questions for which they have been seeking solutions. Some companies came to that realization quite a while ago, and now work alongside academic researchers to give themselves a valuable competitive edge! In fact, it’s now standard practice in the USA.

Which areas are you currently working in?

I’m currently involved in modeling web user behavior using data such as conversations between users in forums, blogs and on social networking sites like Facebook and Twitter. I use IBM SPSS text mining tools to analyze these “conversation markets.” When you start to analyze today’s online behavior, you tend to gather an enormous amount of unstructured data. This approach to gaining a clearer understanding of customers is very different from the data gathered using traditional techniques like questionnaires and surveys.

The challenge lies in being able to analyze the maximum amount of information (the amount of data gathered online is enormous these days), contextualize it, categorize it and analyze it using a simple user interface. If you don’t analyze the way in which web users now discuss brands in online worlds other than those operated by the companies that own the brands (blogs, forums, social networking sites, etc.), you risk overlooking a negative buzz that could be highly prejudicial to the brand, or failing to anticipate the real needs of the market. There are still many sources of data that aren’t properly leveraged. Nevertheless, analyzing and modeling data effectively, using powerful, tailored tools, is what delivers greater efficiency and profitability!

About IBM Business Analytics
IBM Business Analytics software delivers complete, consistent and accurate information that decision-makers trust to improve business performance. A comprehensive portfolio of business intelligence, predictive analytics, financial performance and strategy management, and analytic applications provides clear, immediate and actionable insights into current performance and the ability to predict future outcomes.

Combined with rich industry solutions, proven practices and professional services, organisations of every size can drive the highest productivity, confidently automate decisions and deliver better results.