FacebookTwitterLinkedIn

Originally posted to WWU Münster News by Raoul Volker Kübler.

In his recent publication: Metrics gone wrong: What managers can learn from the 2016 presidential elections, MCM scholar Raoul Kübler discusses together with his colleague Distinguished Professor of Marketing Koen Pauwels how relying on biased market intelligence and targeting the wrong KPIs had fatal consequences for Hillary Clinton’s presidential campaign.

The two marketing scholars argue that focusing on traditional polling measures, which only inquired the preference for a candidate but did not take the intention to vote into consideration, substantially mislead the campaign management. While these traditional polls comforted managers to believe that disinformation campaigns and rumors would not hurt Clinton, probabilistic polls as well as other metrics indicated the opposite.

Both authors continue to point out, that data rich environments – such as political campaigns – do not only provide opportunities but also serious risks for managers. Data can become bone and bane, as managers face more and more complex decision-making contexts in which identifying the right KPIs is challenging. 

As a consequence, managers too often stop information seeking, once a data source helps to confirm a specific believe, without testing for further reliability and validity. This – as underlined by the 2016 Clinton campaign – may however have fatal consequences. 

Both experts therefore suggest to benefit from data rich environments by challenging data sources and comparing effects across variables from different sources. They suggest that Clinton’s campaign executives could have relied on user generated data to confirm or discredit the many polls that saw her in a comfortable lead. Discussions on social media in fact early on showed that Trump had more support than indicated by most polling institutions. Such discrepancies could have then alerted managers of the biased polling results.

More important, the problem of unreliable KPIs is not limited to a political context, but generalizes to the business world. Often marketing managers face similar problems, having to choose between different data streams and sources, when developing strategies and taking complex and lasting decisions such as segment choices or brand positionings.

In their NIM article both authors develop a framework that shall help decision makers to not fall for false information and to identify reliable and valid KPIs. They suggest that marketing managers need to base their choice on a wholistic perspective as well as a critical mind when considering and evaluating data sources. They further point out that marketing theory is a helpful guideline. It offers detailed knowledge about how specific variables and data sources relate to each other and how different variables affect each other. Thus, one can use this knowledge as a litmus test for new data sources. Only if new data confirms these existing insights, one may trust a new source of data and incorporate it into the own analytical environment.

In their article the two authors furthermore suggest different data sources and provide key questions for managers to check and develop KPIs. 

The article is published in the Nuremberg Institute for Market Decisions (NIM – formerly known as GfK Verein) flagship journal “Marketing Intelligence Review (MIR). MIR issues are well known for gathering leading experts in the field to provide latest intelligence and insights from their research. The current issue “Dark Sides of Digital Marketing” is co-edited by Professor Caroline Wiertz from the Thomas Bayes Business School of the City University of London and features articles from leading and renowned marketing scholars such as Anja Lambrecht (London Business School), Klaus Wertenbroch (INSEAD), Catherine Tucker (MIT), Douglas Rushkoff (Queens College New York), and Felipe Thomaz (University of Oxford).

MIR is an open access publication with articles in German and English. View the current issue of MIR.


For more, visit WWU Münster’s News Post.