Personality-based marketing: communicating with customers as individuals through AI

The Head of Psychology at AI startup DataSine, Jergan Callebaut, tells us how to combine psychology and customer data to make their marketing more effective


You’ve no doubt seen ads and received emails for certain products and thought, “That’s useful, I actually really want that…”, or been similarly amazed when you’re genuinely interested in your suggested content on social media.  Far from feeling pushy or annoying, we’re usually happy when it seems like someone has taken the time to understand us and our needs.

But how often does this actually happen?

Whilst marketers can get a rough idea of the content and products that will be most useful and engaging to individuals, this outline is built on basic demographics which are only surface deep.  Things like purchase history, digital footprint, age and gender might give us correlation, but not causation. They show us how the individual has arrived at their destination but give us absolutely no insight into how or why they actually got there.

This is a problem because it treats individual users like mere statistics, which are then analyzed in a spreadsheet, giving no real thought to the individuality and personality behind the outcomes of their unique decisions.

So, is there a way to better understand what actually drives their decisions?

And if we did, could we improve their online experience further so that they receive communications personalized for them?

In a nutshell, yes.

This is exactly what personality-based marketing does. Put simply, it delivers communications to a customer based on their personality. This results in content that the customer wants to see, rather than content they have to see

Most marketing messages – even if targeted based on certain demographics – are delivered in exactly the same way to various different people.  We would never do this in real life when speaking to people we know, so given the option… why would we continue to do this with our customers?

Our personalities underpin how we think, feel and, most importantly, act. By tailoring communications to customers’ personalities, we can increase engagement and therefore reduce the need to send multiple emails i.e. less spam. By using personalized images, words and colors that really resonate with the individual and focusing on features they will care about, we can build a more engaging user experience for every unique customer.  Over the long-term, we can also build far better relationships by connecting with them on a more genuine level, meaning we continue to understand them more deeply and can provide them with better and more useful products.

However, to do this we need to know the personality of the user and how they prefer to be communicated with. Historically, this hasn’t been easy to do for even a handful of customers, let alone thousands or even millions. Fortunately, advances in AI have now made this possible at scale by utilizing algorithms that can effectively model underlying personalities that drive a host of different decisions.

But before looking at the results this technology can give us, how exactly do we define a customer’s personality and more importantly, how can it help organizations?

Powerful personality models

Personality models are a scientific method used to segment customers in a way that enables us to effectively predict their behavior. Through this method, we can leverage decades of impartial psychological research to build up a better picture of why people behave as they do.

Rather than simply dividing your client base into groups with similar demographics and behavior – age, gender, buying history and so on – personality models can help you actually understand deeper motivations.

For example, figuring out what motivates urban-dwelling professional women in their 30s to purchase your products and services requires careful research. Whilst this might work well in certain cases, it doesn’t scale well (or fast) as you add new groups or new products.

This differs from a traditional approach in that different personalities respond to stimuli in unique ways. By studying these responses, we have been able to build up an understanding of how appealing different elements – words, images, colors, fonts, layouts – will be to each personality. This allows us to segment customers much more effectively, before tailoring content and marketing messages to their unique personality profile.

However, the effectiveness of this method depends on the model of personality you use. At DataSine, we rely on the widely accepted Big Five, which has been statistically validated using a technique called factor analysis, rather than being based on the opinions of a handful of people and placed into a theoretical model.

How the process works

The first step towards understanding a customer’s personality is looking at data that demonstrates decision- making, such as:

  • Online behavior
  • Purchasing decisions
  • Engagement with marketing campaigns, on social media, etc.

These decisions can then be used to inform an algorithm which associates them with personality traits. To give a very simple example, through our extensive research we know that people who spend more money on restaurants and bars are more likely to be extraverts, while people that spend more money in bookshops are more likely to be introverts.

We have also built an accurate understanding of content appeal into this algorithm – for instance, that extraverts are more likely to prefer bright images containing lots of people. This is done by asking different individuals to rate the appeal of a variety of content (words, images, colors etc.) and linking this appeal to their personality.

Using this algorithm, we can analyze transactional and behavioral data, score each customer against the Big Five dimensions and predict in advance exactly how appealing content will be to them. It is this ability to predict that gives personality-based marketing an advantage over tools such as A/B and multivariate testing. While these are great for optimization, they can’t predict the future behavior of users.

Personality-based marketing in practice: DataBank case study

With retail banking clients, the information used to build personality profiles is typically about where customers choose to spend their money. In this example with DataBank (name changed for privacy purposes), we used one years’ worth of debit card and credit card spending data for 500k customers.

Once the personality profiles of every customer had been determined, the customer base was segmented according to their strongest traits. In the case of DataBank, we focused on six personality segments:

  • Extraverts
  • Introverts
  • Agreeable
  • Competitive
  • Open-minded
  • Traditional

 

You may have noticed six traits above when we mentioned that we used the Big Five earlier. This is because the Big Five traits each exist along a continuum, so on the “extraverted” trait, we also have “introverts”; “competitive” is on the same continuum as “agreeable” and the same with “open-minded” and “traditional”.

Using this deeper understanding of DataBank’s customers, we tailored content to deliver a personalized user experience to each of them. Many types and elements of content can be customized according to customer personality, including:

  • Images
  • Layout
  • Colors
  • Copy
  • Text messages
  • Call scripts and more…

 

In this instance, we were focused on personalizing images and copy in marketing emails.


The process

Create variations based on target personality segments

As we were using six different personality segments in this campaign, we needed to create six new variations of this the desired email campaign.

Personalize images

The header image of the original email is analyzed and tagged. In this case, with things like ‘travel’, ‘airplane’, and ‘airport’. Similar images that would offer greater appeal to a particular chosen personality segment are offered.

Personalize copy

Text can be personalized at the level of the individual words (e.g. changing ‘wonderful’ to ‘enticing’) and sentences (e.g. focusing on the scarcity of the product rather than the quality), again according to what appeals most to each identified personality segment.

Results

In the case of DataBank the main KPIs were click rates and conversion rates (i.e. the percentage of customers that completed the credit card sign up form after clicking through to the landing page).

Using personalized messaging tailored to the personality segments, we were able to increase click-through rates by 59% and conversion rates by 22%. And, because our algorithms use AI, these numbers are likely to increase with time as the algorithms become more and more adept at understanding the drivers of individual personalities.


Are we really surprised by this?

These results are not so surprising when you consider that currently, the typical marketing email conveys the same message to very different people.

Whilst we do not need to change the overall strategic message of a campaign, or the tone of a brand to suit everyone, we can present these things to our individual customers far more effectively using personality-based marketing.

Using a hybrid approach that utilizes AI while retaining the strategic and creative capabilities of marketers, we get the best of both worlds and can treat every customer as an individual and see the increases in engagement and sales to match.

White papers

From organizing data to AI technology, gain a better understanding of key terms related to the coming AI revolution.

Download

Column Title

  • Link 1
  • Link 2

Column Title

  • Link 1
  • Link 2

Column Title

  • Link 1
  • Link 2