One of the most troubling insights from recent psychological research is that we might not really understand why customers make low-consideration shopping decisions. First, we take a decision, then we invent the reasoning retrospectively. So, how can we truly unearth what consumers want if we can’t really trust what they tell us? This week, our lead Data Scientist Robin Gower explains how Vypr can unlock the difference between stated attitudes and actual purchase behaviour.

Daniel Kahneman describes a battle in our brains between the fast-thinking System 1 (which is quick to act and judge) and the slower, more deliberative System 2. We can take an immediate decision with System 1 and then – when we’re asked to explain our reasoning (let’s say in a focus group or in a supermarket exit survey) – invoke System 2 to invent a rationale after the fact.

This undermines our attempts to uncover insights from consumers. It casts doubt on the validity and usefulness of focus groups. It also affects non-traditional research tools like Vypr. We can’t ask people “would you buy this healthier alternative?”. We may as well be asking “would you prefer to be healthy or unhealthy?”. The responses reflect our personal desires to be fit and healthy (or our recognition that this is socially desirable); not what we choose to put in our shopping baskets.

In reality, we eat a lot of food that is less healthy than we’d care to admit. If you ask someone to account for their choices at the checkout, they’ll perhaps come up with a reason such as “I went to the gym this morning, so I’ve earned these chips”, but this isn’t necessarily what they were thinking as they walked through the supermarket aisles.

So, what’s the solution? How can we tap into System 1 – sometimes impulsive – shopping behaviour? In short, we need to learn from decisions, not opinions. We need to ask our panellists to choose between products (e.g. chips versus boiled potatoes), not concepts. The options we present to the panel should be products that exhibit the attributes you want to compare.

For instance, if I ask a consumer “how important is the tractor mark to your decision to buy milk?”, we’re asking them to engage their System 2 process to invest far more time than they would normally spend on a decision at fixture. Whereas, if I ask two comparable audiences to make a purchase decision on a carton of milk – one of which containing the tractor mark – we can immediately see what impact the tractor mark has on purchase intent. Vypr provides a “Split Test” for exactly this purpose. The difference may seem subtle, but adopting this approach leads to much more robust analysis than would be possible from traditional focus groups or exit surveys.


Sign up to receive news & updates to your inbox