What Associate Editors Look For: Validity, Contribution, and the Future of Marketing Modeling

Written by Veronica Burbulea, Ph.D. candidate at the University of Groningen (The Netherlands)

Marketing researchers have access to more data, more sophisticated models, and more powerful tools than ever before. Yet, according to Aurélie Lemmens, this progress also brings a fundamental challenge: as models become better at finding patterns, researchers must be careful not to mistake those patterns for insight.

In a conversation that ranged from the review process to causal machine learning and artificial intelligence, Aurélie Lemmens reflected on what she has learned from serving as an Associate Editor (AE) at IJRM, and why validity remains the foundation of good research.

Aurélie is a Professor of Customer Analytics at the Rotterdam School of Management, Erasmus University (The Netherlands), and the academic director of the Expert Practice on Customer Analytics at the Erasmus Center for Data Analytics. She has vast experience as a reviewer, as she is on the editorial board of several top marketing journals and has been an AE at IJRM for two years.

The Associate Editor as a Synthesizer

Aurélie describes the AE role as fundamentally constructive. Rather than acting as a gatekeeper, an AE helps authors find a path forward.

“The goal is not to reject the paper. The goal is to try to help the authors find a way to advance the field.”

- Aurélie Lemmens

A key part of this role is to bring judgment and synthesis. Reviewers often approach a manuscript from different perspectives, emphasizing different strengths and weaknesses. The AE’s task is to identify what is truly essential, distinguish critical concerns from optional suggestions, and translate reviewer feedback into a roadmap that authors can use to improve their work.

This process begins with one of the most important responsibilities of an AE: selecting reviewers. At IJRM, AEs choose reviewers themselves, making it essential to find scholars with the methodological and substantive expertise appropriate to the submitted manuscript. According to Aurélie, this step is a substantial, time-consuming portion of the review process, but it can really make a big difference.

What Makes Research Convincing?

When discussing what she looks for as an AE, Aurélie repeatedly returned to one concept: impact and validity. As marketing scholars, we want our research to be useful for external stakeholders. It comes from studying research questions that can change marketing practice or research. Importantly, it also implies that our claims should match what our data can truly support.

“If you have correlational evidence, but then you interpret it as a causal finding, that’s going to be a big problem.”

- Aurélie Lemmens

This concern extends beyond methodological rigor. For Aurélie, a strong paper must also offer a meaningful contribution, either by advancing scientific knowledge or by helping practitioners address important challenges. Thus, her advice for authors applies long before the review stage. Rather than starting with a method or a dataset, she encourages researchers to begin their research with real-world problems and conversations with stakeholders.

“If you talk to companies and you cannot find a relevant question, maybe it means that you didn’t listen properly.”

- Aurélie Lemmens

These conversations, she argues, often reveal questions that are both practically relevant and academically interesting, but it is our task as academics to make the bridge between universities and the outside world.

The Promise and Risks of Modern Modeling

The same concern for validity appears in Aurélie’s views on modeling research. She is particularly excited about developments that combine machine learning with causal inference. Machine learning methods allow computers to learn patterns from data and make predictions without being explicitly programmed for every situation. Causal machine learning and double machine learning allow researchers to leverage large, complex datasets while maintaining a focus on cause-and-effect relationships. Specifically, double machine learning helps isolate the effect of a specific variable on an outcome while accounting for a large number of potentially confounding variables and discovering meaningful heterogeneity.

However, these methods also introduce new challenges. As models become better at identifying patterns and heterogeneity, researchers face a growing risk of mistaking statistical patterns for meaningful insights. A model may identify a demographic characteristic associated with a particular outcome, even when the true causal driver lies elsewhere. As Aurélie noted, sophisticated methods can become “a dangerous weapon” when researchers fail to validate their findings properly. Consequently, procedures such as Chernozhukov et al.’s Best Linear Predictor (BLP) test should become standard practice to make sure heterogeneity is real and not an artefact of the model. The challenge is therefore not simply finding patterns, but determining whether those patterns are causally meaningful.

“With the methods marketing scholars now have at their disposal, the danger is not finding patterns, but mistaking them for insight.”

- Aurélie Lemmens

Research in the Age of AI

Generative AI adds another layer to this discussion. Aurélie sees clear benefits: tasks that previously required substantial time and effort, such as coding in R, Stata or Python, as well as proofreading and editing, can now be completed much more efficiently. At the same time, she worries that researchers may become overly dependent on these tools.

“If you don’t know how to code, are you going to be able to detect the mistakes? Probably not.”

- Aurélie Lemmens

For her, the central challenge is not whether researchers should use AI, they inevitably will, but how they can continue developing the skills needed to critically evaluate its outputs.

Whether discussing peer review, machine learning, or AI, Aurélie repeatedly returns to the same principle: researchers must remain honest about what their methods can and cannot tell them. In a world increasingly filled with black-box tools, that commitment to validity may be more important than ever.

---

Meet Aurélie Lemmens

Full professor of Customer Analytics at the Rotterdam School of Management, Erasmus University, The Netherlands

If you would not be a marketing researcher, what would you be?

“When I was 18, I seriously considered becoming a theatre actor in Paris. I had been involved in amateur theatre since I was very young and absolutely loved it. In the end, I chose a different path and continued my studies instead. Looking back, I still think some of that passion found its way into academia. Teaching, after all, is a little bit like theatre: you are on a stage, communicating ideas and trying to keep people engaged. So, while I left theatre behind, some of those skills are still part of my everyday work.”

What is the best advice you have ever received, and how has it influenced your career or life?

“One of the best pieces of advice I received came from my PhD advisor, Christophe Croux: always remain open to other fields. Throughout my career, I have learned a tremendous amount from reading research outside marketing, whether in computer science, biostatistics, or other disciplines. Many of the ideas that influenced my work originated outside my own field. I still enjoy attending seminars far beyond my area of expertise, because some of the most valuable insights come from perspectives you were not actively looking for.”

---

The article was written by

Veronica Burbulea

Ph.D. candidate at the University of Groningen (The Netherlands)