When companies bring me in to review or teach neuro-based research methods, I can usually predict where the conversation will go. We’ll talk about EEG. GSR. Maybe eye tracking. Often the Implicit Association Test (IAT). There’s usually excitement about the potential of these tools. And rightly so, they can provide valuable insights when used well.

But there's one message I find myself repeating in every session, like a drum I can’t stop beating: humans are complicated.

At this year’s AChemS meeting, I gave a talk on AI in sensory research. One of the demonstrations I gave was intentionally provocative: I asked ChatGPT to simulate a consumer research study using MaxDiff and Implicit Association Testing (IAT) to explore the perception of "freshness" in home fragrance products.

Let me be clear: the goal was not to fake data but to stress-test a study design. By simulating how people might respond, I wanted to explore gaps, assumptions, and how well our methods differentiated between products. Consider it a pilot study by proxy.

I've been asked by at least five different people this week for book recommendations to learn neuroscience, especially from those working in marketing, UX, product development, and consumer research. And I totally get it, the brain IS fascinating. The idea of tapping into real reactions, getting “under the hood” of decision-making, “reading minds”, and predicting behavior is understandably appealing.

When neuroscience intersects with consumer research, it promises an enticing frontier: uncovering what truly drives human behavior, often at a subconscious level. Yet as appealing as it sounds, this approach can run into a significant methodological pitfall—reverse inference.