Psychographic profiling has gained much popularity – and bad press – since Cambridge Analytica helped Donald Trump to become the 45th President of the United States. For digital marketers though, with the reign of social platforms and the rise of big data, psychographics represents an exceptional source of consumer insights.
In the 1960s, the convergence of studies in the fields of personality inventories and motivation research has seen the emergence of a new methodology that quietly grew popular among North American marketers. Almost sixty years later, since March 2018 and the Cambridge Analytica scandal, Psychographics became much more popular, mostly because of bad PR.
However, despite a negative reputation, psychographics used for digital marketing can still be an efficient approach for marketers to gain more significant insights into their audiences and markets. Currently, the most advanced digital marketing agencies are using psychographics to understand their targets better and address them with messages that resonate better with their inner motivations and state of mind.
Furthermore, the hype around psychographics may not disappear anytime soon as recent research has discovered three significant findings that should delight marketing professionals:
First, thanks to their social media footprints, simple statistical models can reveal users’ psycho-behaviourial traits.
Second, artificial intelligence performs better than humans when it comes to assessing human personality (300 Facebook likes are enough for the machine to know you better than your spouse).
Finally, ads targeted through psychographic profiling have better results than classical targeting (i.e., socio-demographics or geographics).
Altogether, these findings should be sufficient reasons to attract a lot of attention from citizens, researchers, marketers, and brands.
What is Psychographics?
While traditional segmentation is based on socio-demographic dimensions (i.e., age, gender, marital status, household income) researchers and practitioners, have projected extra psychological and behaviourial dimensions over samples of populations to capture some essential psychological traits, and predict their decisions about a product (or a political candidate).
With the Psychographics approach, the collection of data, following quantitative means (precoded, objective questionnaires, administered through surveys), leads to the classification of individuals according to their tendencies and behaviours. That means that these 5min quizzes from Glamour or Facebook can reveal much more than one could think about her customer state of mind and buying predispositions.
Precisely, the infamous data scientists from Cambridge Analytica used a simple online quiz, initially based on a personality test from the University of Cambridge (see figure below), to flag respondents along dimensions of the Big Five Model – also known as the OCEAN (Openness, Conscientiousness, Extraversion, Agreeableness, Natural reaction or Neuroticism) Model – and predicted their reaction to political advertising.
The answers to elementary questions were enough to assess someone’s personality across the 5 dimensions of the OCEAN model. During the 2016 US presidential election, the consulting firm used the results of this survey to target voters with high scores of Neuroticism and pushed them specific messages on Facebook.
Source: Grégory Dominé
Using the OCEAN model is not mandatory to leverage the power of Psychographics segmentation for targeted advertising: many other psychographic models can be applied to your segmentation without all the bad buzz associated with Cambridge Analytica.
For example, it is common practice for headhunters and large corporations to assess candidates via psychographic tests such as MBTI (the Myers Briggs Type Indicator is a psychographic model of managerial attitudes and interpersonal abilities of individuals) and PAPI (The Personality And Preference Inventory).
Several other modelisations exist – proprietary or open source – and allow the inventory of individuals and consumers across groups of common psychographic profiles: VALS (for VAlues and LifeStyles), Hofstede’s cultural dimensions, NEO PI-R (Revised NEO Personality Inventory, which is an extended version of the Big Five model) or QOL (Quality of Life, elaborated by the World Health Organisation).
Psychographics & User Data
A promising alternative to psychographic modelisation through surveys is the gathering of enough data to perform an analysis. In the era of Big Data, the number of data sources soared, and some have been tempted to combine them with psychographic methods.
In 2013, Cambridge University PhD Michal Kosinski was the first to reveal the link between digital signals and psychological traits of internet users: likes from your Facebook profile can predict your age, gender, political and religious views, relationship status and sexual orientation (later this year, Cambridge Analytica combined answers from the Discover my Profile’s online survey with Likes analysis of Facebook users to complete their psychographic profiling).
Now, think about the quantity of user data produced every day and provided – for free, quite often – by the social platforms through APIs: Google, YouTube, Twitter, Linkedin, Instagram, Pinterest.
With Personality Insights, IBM used Watson – the firm’s Artificial Intelligence – to extract semantics from Twitter and predict psychographic profiling based on 280 characters-tweets feeds. IBM’s technology is also made available for data scientists and marketers through an API.
Source: IBM Watson Developer Cloud
But don’t get too excited just yet: Facebook’s disastrous management of the Cambridge Analytica scandal has led the public opinion (and the justice) to scrutinise methods associated with psychographics and ethics behind the Big Data buzzword.
Meanwhile, social platforms that built their success over the collection of gigantic amounts of users’ data have limited their access and marketers are held responsible of any misuse: Twitter API general conditions grant permission to mine tweets but explicitly forbids to use them for psychographic profiling.
In Europe the GDRP regulation has described the limits in the use of personal data and made more difficult for data scientists to consolidate psychographic insights into real life ad segmentation.
However, any brand with a Facebook page and the skills required to mine and interpret data can still uncover a lot about their consumers’ behaviours and psychology, even anonymised. Psychographic profiling is not going away.
Grégory is a student at Grenoble Ecole de Management and editor of Content Marketer, a French blog dedicated to content marketing
Photo credits: GD