With the growing ability to capture more data on what consumers do than ever before, it comes as no surprise that Big Data has received a lot of attention. But what has been missing from these discussions is the meaning of the data gathered.
Traditional market research focuses on what consumers say. While observational, ethnographic, and other methodologies that avoid direct questioning have served us for decades, they represent only a small share of the research dollars spent. The dominant share is spent on studies that rely – in one way or another – on questioning consumers about their attitudes, beliefs, intentions, degree of satisfaction, recall, or reactions to new products, packaging, or ad concepts.
As you know, neuroscience tells us that we have two parallel systems in our mind: one designed to get us to ‘do’ and the other to ‘think’. Traditional market research relies largely on rationalizations the consumer’s conscious mind conjures up to make sense of their world. But what drives their purchase behavior is their nonconscious mind, which cannot rationalise, can’t making considered choices, and can’t plan ahead. This means that consumers don’t know why they do what they do, nor can they predict with any accuracy what they will do in the future (except in case of habitual buying) because what drives their actions resides largely in their nonconscious mind.
This is not a particularly new revelation. We have known for a long time that tracking the effectiveness of a campaign on the basis of recall is at best useless and at worst misleading; that group discussions deliver opinions but rarely insights into the consumer’s purchase drivers; that consumers don’t think about a set of attributes when they select a brand, yet are willing to rationalize their actions when confronted with a question asking them to rate brands on a set of attributes.
But what are we to do?
While neuromarketing – the application of neuroscience insights to developing and executing marketing and communications strategies – has developed in leaps and bounds, there has always been a barrier in the research area. Sure, we can test ads or package designs in a neuroscience lab but, while this may help to improve an element of the execution, it’s not going to help the marketer to make the underlying strategy more effective. And, as we’ve already discussed, directly questioning consumers makes no sense, given that some 85% of new products fail, and the majority of these have been tested in group discussions and interviews.
So how do we get meaningful and reliable data about what is driving the consumer’s purchase decisions? Big Data comes to the aid of neuromarketing strategists. It seems incongruous that Big Data and its focus on the behavior of the masses could complement neuroscience, which is concerned with the individual and (largely) nonconscious processes, but it is indeed a marriage made in heaven. Not least because neuromarketing addresses one of the biggest question marks hanging over Big Data: What do I do with all this information beyond planning exposures at the right place with the right message? How do I know what’s truly important? What am I looking for? The answer from neuromarketing is clear: identify market segments based on the goals they are seeking to address with their behavioral choices.
Thanks to Big Data we know more about what consumers do than ever before, while neuroscience research tells us that the consumer’s purchase behavior reflects their attempts to satisfy specific, often higher level, nonconscious goals. It tells us that consumers don’t buy brands because they like them, but because these brands are believed to address their goals. Big Data lets us get serious about segmenting on the basis of goals. As consumers are generally not aware of their higher level nonconscious goals, there’s no point in asking them, but we can often infer these goals by looking at consumers’ behavior.
A simple example: If a consumer wants adventure in their life, then their behavior will reflect this. They may take adventure holidays, prefer adventure movies, spend their weekend abseiling, watch adventure TV series, choose outdoor brand casual clothing, and so forth. Naturally, adventure won’t be reflected in everything they do, as money, time and the need to compromise with companions will occasionally lead to different choices but, given a diverse data set, we should be able to classify this person as an adventurer without having to ask them a single question.
We could call this approach to segmentation ‘implicit’. It is a bit like looking out the window and seeing the trees sway – and inferring that it is a windy day. Implicit Goal Segmentation infers the drivers of behaviour from the actual behaviour displayed.
In collaboration with a global communications group, we started with a massive databank comprising data on more than 60,000 consumers. We extracted the behavioural data and used multivariate analyses to identify segments that show similar behavioural patterns. We then interpreted these segments by inferring the underlying drivers.
For example, we found several segments that are clearly driven by dopamine, but their behaviour is quite different due to their distinct ways of getting their dopamine hits. The segments are unusual by traditional standards, but are directly linked to purchase drivers. They included segments such as
- Dopamine Hunters, External Locus of Control
- Dopamine Hunters, Context Aware
- Socially Retarded, Life is a Video
- Socially Retarded, My Home is my World.
We then developed marketing strategies based on our segmentation for a key product category to test to what extent we could break through category conventions based on the insights the Implicit Goal Segmentation delivered. The results were judged as promising by industry veterans deeply familiar with the product category.
Why am I telling you all this?
Segmentation needs to be an exercise in trial-and-error. There are infinite ways of segmenting a market. Marketers need to try different approaches to segmentation to find a way of segmenting the market that creates meaningful and actionable segments their brands can target effectively.
Implicit Goal Segmentation represents a great starting point as it does not rely on rationalizations: when we ask consumers about their likes, needs, attitudes, reasons for purchase, intention, and preferences, we simply get rationalizations. The consumer does not know what is going on in their nonconscious mind and thus cannot tell us. Implicit Goal Segmentation avoids this pitfall. Another advantage is that Implicit Goal Segmentation is based on a massive database, avoiding the pitfalls of small samples used by most neuroscience-based market research techniques, which are unlikely to be representative of the market.
More work needs to be done to explore the Implicit Goal Segmentation approach in-depth and assess its applicability for different categories. For example, Implicit Goal Segmentation may be unsuitable when dealing with a category dominated by habitual purchases as our focus on purchase drivers is likely to fail. Maybe you have an opportunity to test-drive implicit segmentation? If you do, let us know how well this approach has worked…