Joel Rubinson on making 'calorie-poor' research nutritionally valuable

We were joined on our podcast by Joel Rubinson, the president and founder of Rubinson Partners, a marketing and research consulting firm for a ‘brave new world’. Joel is the former Chief Research Officer of the Advertising Research Foundation and is a member of the Faculty of NYU Stern School of Business where he created the first graduate level business course, social media strategy. Joel was also a baseball pitcher at NYU and is an accomplished blues harmonica player as well. We started out our conversation talking a little bit about Joel’s interesting background and recent market research projects, including some innovative programs he has developed to help marketers quantify the value of brand building and how to best target people with advertising - essentially using research to boost ad effectiveness.

During the podcast, he focused on how important it is in your research strategy to realize that insights alone are not enough, and that every insight that you're offering has an embedded prediction in it. He encourages researchers to flesh out what that embedded prediction is “because then you're really directing action and you're also setting the stage for validating whether the insight was true or not. So for me, insights that aren't accompanied with predictions and indicated actions are kind of empty calories.” He said there are a couple different kinds of predictions to pursue, depending on goals.

Joel goes on to explain some of the design of the research he’s doing in more detail, and the history behind it. “It was really an idea that I had, that had been cooking in me for years - that the probability that somebody had to buy a brand, obviously, affects their likelihood of purchase. My thought was it might also affect their responsiveness to advertising over and above their baseline probability of purchase. So I wanted to examine that question. Was there a mathematical linkage that we could infer and prove between someone's baseline probability of purchase and how they respond to advertising? So we started with that…We brought in academics to review the methodology, from Oxford and UCLA. So we knew it was pretty bulletproof.”

We discuss some examples of Joel’s methodology at work, and what the next steps are after the insights are delivered. From integrating multiple shopper data points, such as detailed purchase data, credit card data and even location data, more targeted ads can be served to the right audiences. He says that based on the data: “you should have differentiated marketing strategies. It's not just about sending the right message to the right ID. It's about being selective in your ad serving. I've actually built an audience optimizer where the optimal levels of spending from one audience to another can be 250 times.”

After covering some real-world examples with well-known brands and retailers, Joel talks about various metrics, such as KPIs and even NPS (which he maintains is not a very accurate measurement, need to be aligned with an underlying model in order to be meaningful. Proper analytics and modeling of consumer processes, along with understanding advertising's relationship to growth, lead to better metrics. Joel advises researchers to consider the implications of their insights, connecting them to future predictions and actionable marketing decisions. 

Want to learn more about this innovative approach, the “movable middle” for advertising and more? Listen to the full episode!