By Nele Van der Elst and Lieven Buellens, Rogil
On December 13th 2012, Rogil attended the annual Baqmar conference where we witnessed a very interesting and inspiring afternoon and evening program with top presentations. We now experienced that research is lifting to the next level.
The datamining track opened with John Crombez (Secretary of State for the Fight against Social and Fiscal Fraud) who explained how the Belgian government tries to lower tax fraud in Belgium through data mining. Belgium has a very high income tax rate, however tax income per person is very low. The Belgian government realized that data mining is useless when working from an ivory tower. The involvement and feedback of the field force regarding selection criteria used in data mining and the internalization of external knowledge regarding data mining are crucial in increasing the hit ratio of fraud inspections.
Soroosh Nalchigar (University of Toronto) spoke about the use of online search behavior to predict future purchases. Based on a clear segmentation and a lengthy investigation of online behavior of respondents using Yahoo Toolbar, he was able to predict future online sales more accurately.
Daniel Soto Zeevaert (Deloitte) passionately talked about the use and utility of conjoint analysis to get a better perception of the drivers of value and thus value itself. He clarified that conjoint analysis should always be used regarding specific segments. Every segment can have different drivers of value and by identifying these drivers there it is possible that marketing campaigns can be better adjusted on the needs of consumers and thus resulting in a higher impact than when targeting the entire market.
Normally the consumption pattern of consumers online is investigated by certain flagpoints in their online behavior. Amit Phansalkar (Kantar Media Compete) said that most companies take into account typical online search patterns like for example looking at product reviews online, comparing multiple products etcetera. Based on these patterns companies can try to be relevant for online consumers on crucial points. However through data mining the consumer habit cycle can be investigated by looking at the clicking pattern of visitors on websites.
The evening program started with a nice story of Yahoo. Laura Chaibi showed us in real time the enormous amount of data which is gathered every second on yahoo. However, this is only giving us information, but not insights. The researcher still has an important role in turning this huge amount of information into insights. Furthermore, it is very important not only to look at the data itself, but also to look for the motivation behind the data. What is the context in which the answers were given? Why are respondents giving these answers? Which emotions are linked to these answers? The researcher is there to make the link between what consumers say and what they eventually do. This by using implicit reaction time measurement, eye-tracking, etc.
Pascal Mignolet (DE Master Blenders 1753) talked about the great innovation lottery. 80-90% of innovations fail. This was 20 years ago a fact but, surprisingly it is still the case today. The value of consumers as a source of information is declining and consumers’ validation becomes less reliable. But also employees become uncomfortable with innovations within their company. Therefore traditional models have to be reviewed. What is the role of the consumer/employee in the innovation process? If we want to decrease the number of failing innovations we have to change the way we research and not only rely on consumers opnions but have to look at research in a different way and evolve to a world: where we ask less and observe more, try to measure as realistic as possible (test a shelf not online but with a real shelf), less interviews more dialogue, less representativeness and more relevancy, more real life less lab. We have to leave our comfort zone, even though it lacks standards.
Ben Smithee (Spych Market Analytics) talked about the connectivity of consumers by their phone. It is the second thing for which you run back in your burning house. Products have to be market through consumers not to consumers; e.g. by using social media, social listening, being with your brand in online conversations. You have to get involved in social media with your brand. The future researcher has to be a Jack of all trades: a true connection and trust with consumers, being in the consumer’s cycle of trust, making a connection between methods and technology.
And in between the speakers we enjoyed a fabulous concert of the first ever Baqmar Band.