We don’t know each other at all, but I suspect you would be able to predict with fairly harrowing accuracy how I might react if a Silverback Gorilla suddenly materialised in my office.
The reason I mentioned this imaginary scenario in which I would, of course, wrestle the Surprise Gorilla away immediately is to illustrate that, in some contexts, almost everybody reacts the same. But I also mention it to illustrate that those situations are a) vanishingly rare, and b) almost always very extreme cases. For almost every other occasion, we need rather a lot of information about both the individual and the situation to offer a truly robust prediction of their behaviour.
As marketers, one of our most basic needs is a way of predicting behaviour — it is a prerequisite to efficiently being able to change it. Historically, we’ve done this by theorising and speculating — from the Edward Bernays’ Freudian theories to Bill Bernbach’s faith in creativity. But in recent decades we’ve got more and more data driven; the only problem being data is only as good as your ability to parse it out, and the underlying assumptions in our inputs have huge effects on their outputs.
What we don’t often do is admit to ourselves how contingent all the factors going into our precious demographic segments really are
In trying to predict behaviour, it has become accepted that understanding consumers’ personalities is key. And that makes sense. So we go to great ends to create complex, richly detailed psychometric ‘profiles’, then use them to segment our audiences. We do this to make our targeted content more effective; to make our targeted content more cost-and-time-efficient; to ensure that our targeted content is actually properly targeted. What we don’t often do is admit to ourselves how contingent all the factors going into our precious demographic segments really are; contingent, primarily, on context.
Person A is ‘interested in cars’, sure. But in what context does that manifest as actual goal-orientated behaviour? Because very few people walk around perpetually interested in anything. In fact, we have a word for that: obsession.
This might seem like needless nit-picking; segmentation is ‘effective enough’. We might not be nailing it 100% of the time, but it’s accurate enough to get the results we need. And that’s the attitude plenty of marketers are taking.
The problem is that it ignores two key issues.
First is the growing issue brands are having with the decrease in audience’s irritation threshold — our recent research at Boostify has re-confirmed what should be incontestable at this point: users will not tolerate too much irrelevant, disruptive targeted content, and being forced to do so will sour them towards any brand and business with gall to subject them to it. However much cache you’ve built up, poorly targeted, poorly timed content will bring it crumbling down in an instant.
So that’s not ideal.
But the second, perhaps more widespread result of under-valuing context, is that it means settling for results you could be far surpassing. If we accept that psychographic profiling has become more and more effective and nuanced over the decades since Arnold Mitchell developed the Values Attitudes and Lifestyles method, it seems strange to decide that it won’t continue to improve, especially as we now have such an awe-inspiring range, depth and breadth of data to work with.
The way it will improve will be embracing a fuller understanding of context.
The Science of Personality
It took a long time and many floundering attempts, but by the early 1960s a personality model — the Big Five Personality Traits — came into general acceptance amongst the academic community. Then came along a man by the name Walter Mischel, who went and complicated the whole thing.
Mischel argued that the typical correlation between personality traits and behaviour was actually rather modest. In fact, he claimed that, in-of-themselves, the traits personality psychologists had been measuring were really only slightly better at predicting behaviour than astrological signs.
Instead of seeing people as made up of a collection of static traits, we need to see personality as a set of unique, individual cognitive and affective variables which help people interpret and make sense of situations
Mischel’s argument was that personality researchers had underestimated the extent to which context shapes people’s behaviour. His alternative — Situationism — came into vogue thereafter, but soon Personality Psychology would reassert itself, and has since regained its place as the accepted measure of behaviour.
However, many people simply leave Mischel’s work there — a relic of the academic tug-of-war. But the fact is this underestimates and undervalues his work. Because Mischel didn’t simply dismiss the existing research outright; rather, he found this flaw in it and proposed an amendment to it. Personality is a good predictor of people’s behaviour, he said — it’s just that it has been conceptualised poorly.
Instead of seeing people as made up of a collection of static traits, we need to see personality as a set of unique, individual cognitive and affective variables which help people interpret and make sense of situations. Individuals have idiosyncratic, habitual ways of interpreting and evaluating different situations and contexts; these interpretations are what truly drive behaviour.
You behave differently in the boardroom than you do getting a coffee, and different around your parents than you do your friends. Try taking two friends from two different periods of your life and asking them to take an online personality test as you. You’ll almost invariably find that their answers diverge greatly. And therefore, that the personality ‘type’ it’s determined that you possess will be at least a little, if not hugely, divergent.
It doesn’t really make sense to talk about personality and context separately.
And that’s because they got to know you in different contexts at different times. You will have had different goals at these different times, and have been surrounded by different stimuli. In fact, all the things a psychographic profile would consist of — interests, attitudes, values, aspirations etc. — are themselves context dependent: they will almost invariably fluctuate between differing context at different times.
All of this ought to be intuitive and obvious: it doesn’t really make sense to talk about personality and context separately. Of course, we all have traits which are consistent across differing contexts. But even a dense, fully realised psychographic profile isn’t going to help us very accurately predict an individual’s behaviour without any context. So why is this fact so routinely ignored by marketers?
The future of CRO
Perhaps it is a bit of a straw man to suggest marketers ignore the relationship between character, context and behaviour. What would be fairer to say is that a) they often focus on one or the other, and b) they’re not well equipped to deal with both at once.
Part of this is just a technical issue: martech to facilitate such a marriage of character and context is severely lacking. And there are simply so many muddy variables to consider, even if they could many marketers would willing to forego the gains to avoid having to trudge through them all.
Part of the problem is also structural: context, when it isn’t an afterthought, is usually taken up by the creative team, whereas character targeting is taken up by media buyers and tech types. So we see some phenomenally inventive uses of context, where content is tailored to and often plays on its placement and format; and we see some extraordinarily accurate characterological targeting, where only exactly the right people see a given message.
But imagine if we put those two teams together, and content was purpose built not just for a specific context with a specific message, but for specific individuals in that context, taking into account their idiosyncratic ways of interpreting and evaluating that context. This is the future of CRO, and businesses who are able to crack it will be so far ahead of the game it’s hard to overstate it.
What they’ll need to remember, however, that even targeting specific personality types in specific contexts at specific times won’t in of itself be enough to guarantee sales. They will need to conscious of the final piece of the puzzle, which we explore in the next post of this series — goal-orientation.