OPINION27 August 2015

Mind measurement


In the latest in Colin Strong’s series rethinking areas of market research, he turns his attention to the importance of attitude measurement and psychometrics.

Market research has historically focused on two broad areas – attitudes and behaviours. Obviously within this sweeping generalisation there lies huge variety – so attitudes can include needs, interests, future intentions and propensities. 

Behaviours cover everything from what was eaten that morning to what was spent at the supermarket.  But what market research has traditionally done much less of is measuring consumers’ psychological structures. We tend to spend little time measuring personality attributes for example. Some may rightly argue with this, as there is a grey area between attitude measurement and psychometrics (measurement of psychological attributes), but, in the main, it’s a fairly peripheral activity. We tend, for example, not to see the psychometrics forming a core part of a U&A or segmentation study.

But this is puzzling, as having this understanding of your customer base is potentially a very useful tool for marketers. So why has there been this reticence? In my mind it’s for a number of reasons. 

First, the dominant mindset in marketing and market research has been a fairly positivist one. We are interested in the particular inputs (attitudes) influence outcomes (behaviours) and less excited about the mediating variables that may shape these attitudes (such as personality). Second, while a wide variety of studies show that personality traits shape consumer decisions, it is less clear how a marketer would use this information to guide their activity and how this varies by context. And finally, psychological measurement is tricky as questionnaires are often long and the nature of the question set could be considered irrelevant and intrusive in a commercial context by those completing them.

Psychological insight

But marketers are starting to see the value of gaining a more psychological insight into consumers. Recent publications have provided momentum on this topic. First, Daniel Kahneman’s book Thinking Fast & Slow brought system one/system two thinking to a much wider audience. But perhaps more importantly, Byron Sharp’s thinking on the way in which memory structures shape How Brands Grow captured the imagination of senior management who have cascaded down the imperative for more implicit explanations of consumer behaviour. 

More recently Persuasion Profiling has offered a guide on how to measure consumers’ propensity to be ‘nudged’ to particular outcomes.

This thinking is part of a wider move to an experience economy where technology and data is rapidly transforming our world away from products and towards services. 

Think Nespresso as a prime example of the ‘servitization’ of what might previously have been a functional sell of a basic product (coffee) which required fairly familiar research around product features, pricing and positioning.  Because services such as Nespresso are technology-enabled, there is a stream of data about user behaviours that brands can increasingly use to, not only understand customer behaviour, but also their psychological profiles, allowing them to optimise the customer experience.

How can market research operate in this changed environment?

Behavioural economics continues to be a source of huge excitement and activity, increasingly used to explore the way in which environmental factors can interact with cognitive biases and implicit attitudes to influence outcomes.  But what has had less focus is the area of psychometrics, despite its very obvious benefits. 

Given changing marketing priorities perhaps it is time to rethink the way in which the market research industry uses psychometrics. Here are my recommendations on how we can respond to the growing level of awareness and interest in the opportunities:

  • Get better at psychometrics development: our industry is actually pretty good at developing usable measurement tools. There is no reason why this cannot be done for measurement of psychological traits – some of us already do this. There is an argument for wider adoption.
  • Showcase how marketers can use psychometric outputs:  we need more use cases on the way in which psychometrics influence consumer behaviour beyond vague generalisations.  Using a Community Panel for a ‘test and learn’ environment means it is possible to explore what works, which helps us to develop guiding principles.
  • Move from questionnaires to data: if we can predict different psychological attributes from data, then it is possible to profile the entire customer base rather than being limited to the survey sample.  This makes it a significantly more powerful set of tools.

But there is a wider ethical debate that the market research industry needs to be leading.  Fundamentally is it right for brands to be using this sort of information to guide consumers in ways that are not explicit? Some would say no, reflected by developments such as darkpatterns.org, which attempts to draw attention to ways in which user interfaces have been “carefully crafted to trick users into doing things”. But perhaps, as Tom Loosemore, of the Government Digital Service recently tweeted (in a personal capacity), “Nothing wrong with defaults, so long as the citizen actually understands consequences of their choice”.  The ethics and practices need to be debated and developed hand-in-hand.

As we move into an environment where much of market research will inevitably be about ‘Humanizing Big Data’ then the opportunities and pitfalls, the risks and rewards of understanding the techno-social aspects of our increasingly data-driven lives, is becoming critical. 


5 years ago

Hi Colin, great article. The debate about the ethics of hidden persuasion has a long history. But I think it is something that has been very much overlooked in the 'excitement' that surrounds behavioural economics. Of course, this isn't just a problem for market research. The tension between system 1 strategies that seek to exploit automatic psychological processes and people's deep desire for individual autonomy is evident in every field where behavioural economics is being applied. Think of the oxymoronic concept of ‘liberal paternalism’ that lies at the heart of Thaler and Sunstein’s ‘Nudge’… Tom Loosemore’s, comment (“Nothing wrong with defaults, so long as the citizen actually understands consequences of their choice”) doesn’t get us very far. Default strategies and transparency don’t really work together. If we tell people about the consequences then it is not a default and it ruins the effect! Marketers, researchers and policy makers need to confront the ethical issues around hidden persuasion head on rather than trying to rebrand it or sidestep the issues with glib references to transparency. My view is that the science has established that hidden persuasion is a simple fact of life. At Nudgestock a few weeks back Richard Thaler commented that Eve was probably the original choice architect! Whether it is designed or not, hidden persuasion can’t actually be avoided. So, for marketers, researchers and policy makers, there is no binary choice between explicit persuasion (good) and implicit persuasion (bad). It is a spectrum. But we do need a more open and informed debate about much we should deliberately exploit the implicit side. Thanks for highlighting this and thanks for the reference to darkpatterns.org.

Like Report

5 years ago

Hi Ian Thanks for your very interesting comments. I think much of this quickly gets to a discussion of what it is to be human. Are we fundamentally creatures that are shaped by choice architecture or does our consciousness mean that we have insight and self determination. Obviously the answer is 'it depends' but this is the bigger debate that MR needs to have. William Davies, among others, has written very persuasively about the way in which we seem to have tipped our scales towards a model where implicit attitudes / cognitive biases are considered more important. I think we are all struggling to know where that line actually is and indeed, the extent to which it is ethical for brands to cross that line.

Like Report