Bring consumer neuroscience to your

Following on from decades of research in human decision making and consumer psychology (Cognitive-Affective-Conative model), we have developed the BRAIN model. It stands for Behaviour, Rational, Affective, Intent, Network. 

Behaviour – we build in category behaviour into the model. Typically this will be the products each consumer buys, considers, and rejects. It also includes their culture / location if needed, so that the predictions made are culture-specific.

Rational appraisal – in our surveys we measure beliefs and explicit evaluations of product or service features and benefits. These are cognitive or rational judgments. They involve reasoning, comparison, and expectation.

Affective appraisal – this is where implicit responses do their work. Implicit reaction time testing works best when it measures feelings: relief, confidence, pride, enjoyment, reduced stress, improved sense of control and so on. Affect is not just about emotions, it’s about value: “Will this make me feel the way I want to feel?”. This is the emotional or experiential payoff that a product or service is aimed at delivering.

Intent – this is the purchase motivation, decision to buy, or actual purchase behaviour. We can measure these through questions about purchase history, purchase intent, willingness to pay, or actual data from purchase receipts.

Network – this refers to the decision-making network or sequence that are relevant for making predictions about consumer behaviour: (1) the consumer has a problem, (2) a product or service proposes a solution to the problem though its features and capabilities, (3) these proposed solutions provide benefits – “promises” to fix the problem the consumer has, (4) the consumer perceives the features and promises and makes a rational appraisal of them, (5) at the same time the consumer has an affective response to the features and the promise, (6) the output of the sequence is an intention or motivation to want the product or service, (7) the consumer decides to make a purchase. 

All of this data are collected form our surveys. They feed into the modelling process. Some elements will be more important and more predictive than others. Some rationalised views may be important for the consumer, but not predictive of their buying behaviour. 

The data are used to train a large neural network. It uses associative or machine learning though thousands of brain-like processing units. It identifies patterns in the data, especially looking at how all of the behavioural, rational, affective appraisals are associated with the final purchase decisions. This gives us the final list of appraisals that predict a purchase. This list is weighted, some are vital others less important, but in combination they all serve to convince consumers to buy or not to buy.

Once we have the model of a specific market category, we use it to simulate any kind of “intervention” by a brand in the category, such as an ad, a marketing concept, a communications strategy, new packaging, a new product concept, a marketing video, taglines, claims, and so on. We do this in our Amethyst Asset Tester. It is multi-model, meaning that it can assess images, text, audio, and moving images, and combinations of these.