Communicators and marketers can now adopt a personalized approach to their work, ideally one based on behavioral science. But the execution lags behind the science while the claims of some marketers as to what personality marketing can do far exceed it. Moreover, public controversies like the Facebook and Cambridge Analytica story threaten personality marketing’s potential before it has really matured.

It’s important not to judge a field by its worst actors. Marketers, communicators, and the public alike deserve a better understanding of personality marketing — what it is, how it works, and why it matters.

The personality targeting controversy

Beyond the allegations of misuse of personal information gleaned from unwitting participants in social media, the Cambridge Analytica controversyraised an aspect of marketing that few people knew much about: the targeting of people based on not only on their past behaviors and explicitly stated preferences, but based on their underlying psychological profiles. Cambridge Analytica claimed to use personality trait science to better tailor messages for its clients.

“This allows us to nuance our messaging,” said the Cambridge Analytica CEO, before an audience of 250 communications executives six weeks before the 2016 US presidential election, “so that rather than serving the same advert to 100 million people… you can sub-segment people by personality and change the creative to resonate with individuals based on how they see the world.”

Beyond traditional personalization based on demographics, or consumer self-expressed desires, this kind of customization claims to interpret basic human drives and match issue messaging with personality traits.

Some view psychometric or personality targeting as a dark art. But as behavioral scientist Cass Sunstein has cautioned,  there are sound uses for personal data on social media if handled ethically. We believe that includes personality marketing. Personality insights and other aspects of behavioral science offer opportunities to better connect with individuals, and if done ethically it can be beneficial for consumers and businesses alike.  Personality marketing can create a better match for products, services or experiences. And in sectors like health care, it could have even more positive affects, with better messaging leads to healthier behaviors.

What is personality science?

The American Psychological Association defines personality as “individual differences in characteristic patterns of thinking, feeling, and behaving.”  Scientists have believed for centuries that humans have a mix of traits that determine the way individuals interpret the world and how they subsequently behave. Studies of identical and fraternal twins suggest that about 40 to 50% of our personalities can be attributed to our genes. Psychologists converged on the current and widely-accepted view of personality science about 25 years ago, simplifying hundreds or even thousands of personality descriptions to just five core factors with supporting facets for each. Called the “Big Five” or “Five Factor,” they form an easy acronym of OCEAN (see sidebar).  An individual can be mapped to a sliding scale from low to high on each factor. Psychologists have run tens of millions of people through the personality tests around the world with high reproducibility and consistency.

For marketers, communicators, and even public health agencies looking to promote healthier behaviors in large populations (diet, nutrition, exercise, quit smoking), the potential payoff of using personality science is to be able to better match how you engage individuals by personality profile, and to predict behaviors by personality traits.  No marketer wants to present a message that is off-key or even irrelevant; personality science offers the chance to empathize with individuals, and engage them with the message, advertisement, or content in a way that is more likely to resonate with them.

How do you test consumers for personality profiles?

Until very recently, the assessment of psychological traits (also known as psychometrics) was almost inseparably tied to questionnaires. Ranging from just 10 to more than 300 questions, these questionnaires ask respondents about the extent to which they agree to statements such as “I am the life of a party” (Extroversion) or “I get chores done right away” (Conscientiousness). While such questionnaires provide researchers with an easy and pragmatic way to measure people’s psychological profiles, they are prone to unintentional misrepresentation, especially in certain contexts outside of the lab. For example, no job candidate in their right mind would indicate “strongly agree” to the statement “I make a mess of things.” Plus, these questionnaires were difficult to scale up beyond a few hundred or thousand test takers. If insights about people’s psychological make-up could only be gleaned at relatively small scale, how could marketers and communicators leverage such insights on millions and millions of potential customers?

Only about five years ago, the newly established field of computational social science provided an answer: digital psychometrics. Instead of relying only on people’s responses to self-reported questionnaires, scientists started using people’s digital footprints—their Facebook Likes, Tweets, browsing histories, and more—to make inferences about their personality (with their consent). Studies compared large groups of individuals’ traditional personality questionnaire responses with those same people’s social media behavior to see if their personalities could be accurately inferred merely by decoding their digital footprints. Based on large datasets containing both people’s responses to traditional psychometric questionnaires and the information captured on their Facebook profiles researchers were able to identify empirical relationships between specific digital footprints and specific psychological traits. For example, certain correlations arose between liking a certain kind of music or food and specific personality traits. The more content, the more accurate the assessment. Using the Facebook Likes of U.S. Americans (around 250 Likes, on average, in 2012), computer algorithms were better at judging a person’s personality than their co-workers, friends, and even family members.

While still in its infancy, there are more and more commercial attempts at predicting personality from people’s digital footprints. IBM’s Watson Personality Insights, for example, uses natural language processing to digest bodies of text written by a specific user, like tweets and blog posts, to unearth their personality traits, needs, and values. With the rise of such services, insights from digital psychometrics will no longer be limited to mainly academic contexts, but will become available to industry at large.

Why use personality marketing?

Again, the theory is that if you can match the tone and framing of the communications or marketing with the personality profiles and thinking styles of potential customers, patients, voters, or those whose behavior you’d like to change, you can boost effectiveness.

For example, look at retail shoppers. Some do not particularly like shopping. They take a functional approach. It’s a chore. They are not wedded to brands and are motivated by price. Psychologists call them “utilitarian” shoppers. Meanwhile, “hedonic” shoppers enjoy shopping and love brands and join loyalty programs. They identify with brands and may use them to signal who they are and what they stand for. They may look identical when viewed through traditional demographic data, but you would be sorely mistaken to treat them the same way. Certain personality profiles correlate with utilitarian versus hedonic shoppers. So, if you can know shoppers’ personalities, you can customize how you engage them.

But these findings are not limited to retail shopping. In fact, tailored communication has proven highly successful in the context of health care and health communication. We know that people show higher compliance rates when receiving messages that are customized to their individual motivations, and we also know that such messages help in changing a number of cancer-related behaviors, including smoking, dieting, exercising, and regular cancer screenings.  What if we could not only increase the chances that a customer buys a handbag, but improve their quality of health or the uptake of flu shots or vaccinations by tailoring the messaging to different personalities and cognitive styles?

But does personality marketing work?

The scientific evidence is consistent and clear: one can increase the effectiveness of marketing messages and other types of persuasive communication by tailoring them to people’s psychological profiles (see compendium of studies here curated by IBM).

The problem is that these results come primarily from the lab. Therefore, the usefulness of these insights for real-life customized marketing remained limited. The lab is not the same as the market, and questionnaires are not the same as personality inferences based on internet data. But there is reason to believe the science will hold up at scale.

The Ogilvy Center for Behavioral Science, where one of us works, has collaborated with Kantar, a global data and research company, to start layering personality traits research over the existing data from large panels of Kantar Lightspeed respondents. The idea is to find new correlations between specific personality profiles and other tastes, preferences, and behaviors. Kantar has also made progress in overcoming the self-report bias in personality questionnaires with a novel approach that forces respondents to prioritize descriptions and to do so in context.

So, what do we know about whether personality marketing works in the real world? One of us, Sandra, pioneered a study to determine whether the application of digital psychometrics to tailored communication could significantly impact the effectiveness of large-scale, real-life advertising campaigns on Facebook. While Facebook does not offer direct personality targeting, it allows marketers to do so indirectly via the option of targeting people based on interests. If liking “Lady Gaga” on Facebook, for example, is associated with the personality trait of extroversion, and liking “Stargate” goes hand in hand with introversion, then targeting users associated with each of these Likes allows one to target extroverted and introverted user segments. (The correlations need to be sufficiently strong, at least when taken together; if Lady Gaga Likers are only barely more extroverted than non-Likers, that’s not very useful.) Of course, the more data inputs such as Likes, the more accurate the analysis.  The results of three campaigns reaching over 3.5 million users suggest that personality-matched advertising creatives significantly outperform their mismatched or neutral counterparts. In other words, in practice, this sort of social media-based personality marketing does appear to work.

However, in light of Cambridge Analytica, and claims that its psychological targeting tipped the election, we caution against any extreme claims related to personality marketing. The early evidence may be promising, but the field is still relatively young.

The ethics of personality marketing

The essentials of gathering and using personality traits ethically should follow the general guidelines of other behavioral science research of consumers, employees or patients. They include: transparency of intent and usage; abiding by privacy laws and regulations; and aligning researcher/marketer interests with those of respondents (in other words, help them rather than exploit them).

That last principle is the right starting point for marketers: is your use of personality research actually making your customers better off, or just helping you? As the field evolves, marketers should look to the research community for inspiration and guidance on transparency. And, of course, businesses must comply with the law.

Putting personality marketing in action

Given the promise and accessibility of this new form of communication, how should marketers get started?

In our experience, the first step is to understand the challenge or goal you’re trying to achieve. Is it to align employees with corporate goals, or to promote smoking cessation, or to increase uptake of vaccinations, or to change consumer behavior, or better segment consumers by what really motivates them (which they cannot articulate)?

Next, identify the cognitive biases and heuristics serve as barriers or drivers along the way to achieving the goal. (The Ogilvy Center for Behavioral Science has built a tool to navigate thousands of studies to surface the relevant biases.) Map the biases to steps along the consumer (or patient) journey. Doing so will help you identify steps along that journey where creative communications or content can help consumers overcome specific biases or other hurdles to a decision or new behavior.

Once you have a strong understanding of the customer journey, you can run a personality test and combine it with other data to reveal correlations between personality traits and certain behaviors, preferences, or mindsets.

The final step — and the “art” of personality marketing — is to craft the messaging, advertising or content to match different personality profiles while also considering the stage of the customer journey at which you plan to engage. This isn’t easy, by any means. But it offers the opportunity to create the most effective and empathetic messaging with different groups of customers.

Personality marketing is just one aspect of a new, fast-emerging approach to understanding people from the inside out. We can now move from observational oddities of what makes humans “Predictably Irrational,” as the behavioral scientist Daniel Ariely has written, to decoding what truly moves individuals — at scale — and engaging them on their terms. How we do this will determine whether it is used for empathetic communication and positive outcomes, or for manipulation and exploitation.

Christopher Graves is the founder and president of the Ogilvy Center for Behavioral Science, a recent Rockefeller Foundation Bellagio Resident, and Chairman Emeritus of Ogilvy Public Relations, as well as immediate past Chair of the PR Council.

Sandra Matz is a computational social scientist and Assistant Professor of Management and Organizational Behavior at Columbia Business School.