Personality & marketing effectiveness. What makes CTR and conversion go through the roof?

After climbing up Schlossberg’s steep and slippery slope in Freiburg, the home town of the University of Cambridge’s computational social scientist Sandra Matz, in early January 2017, we found an excellent place to talk about her research. It applies Big Data analytics and psychological theory to digital marketing. She elaborates on the significant effects personality centric targeting and matching can have on marketing effectiveness. Additionally she reveals how spending money can make you happy. And lastly she talks about the effects personality-centric tageting can have on elections.

Time-sequenced questions and transcript of the interview below.

Questions

• 00M08 Sandra Matz you are a doctoral researcher at the University of Cambridge. Can you give a brief introduction about yourself and your research interest?
• 00M37 Before I go into depth just a few introductory remarks ‘’money can’t buy me love’’ the mega hit of the Beatles, to what extend does it confirm or deny your research on spending and happiness?
• 02M28 Personality centric spending, targeting and visualizing are main fields of your research interests. What made you decide to go into them?
• 03M41You did some research on spending money and well-being. If you look at it, what can be applications of that research and what parties might benefit?
• 05M12 Your research is predominantly done online and moreover in the social media. What does this mean for the application of doing research offline, in the ‘’real’’ world so to speak?
• 06M39 Why is it that personality centric targeting results in these significant increases in CTR (click through rates) but moreover gigantic surge in conversion rates?
• 08M14 What is the difference according to you between consumer centric on the one hand and the personality centric on the other hand in terms of approach? And what do companies need to have in place in order to make this move from customer to personality centricness?
• 09M12 In some of your research we see that you target on one big 5 factor only. In order to improve the effectiveness of personality centric targeting, how many of the big 5 traits should at least be targeted simultaneously?
• 10M34 Let’s assume I were a marketer and I would start with personality centric marketing and I had a limited budget. What would you advise me to spend my money on: On personality centric messages or personality centric product fit?
• 12M16 Already in 2012 your fellow researchers Kosinski and Stillwell showed that personality centric targeting was a next step in marketing. And now your latest research seems to be indicating that the benefits are significant in fact huge. How come, do you think that industries like retail hardly seem to make use of these findings?
• 14M38 Early December 2016 the Swiss magazine ‘Das Magazine’ sparked discussions about personality centric targeting and election results, the American result even. What can be the consequences of this discussions for the application of your research findings and the research itself?
• 16M23 I will try to wrap it up, your fields of interest are personality and well-being, personality and advertising effectiveness and personality and advertising aesthetics. What are the three important and unanswered questions in these fields and what does that means for your research priorities?

Audio

Transcript

00M08 Sandra Matz you are a doctoral researcher at the University of Cambridge. Can you give a brief introduction about yourself and your research interest?

Hi, I am Sandra Matz I am a computational psychologist and consumer researcher. I’m at the University of Cambridge. So what I am interested in is really how we can use a combination of big data and psychological theory to personalize our experience online and the way that we experience the digital world.

00M37 Before I go into depth just a few introductory remarks ‘’money can’t buy me love’’ the mega hit of the Beatles, to what extend does it confirm or deny your research on spending and happiness?

Well, I have to say that I really love the song and I think it very nicely captures the belief that actually money is something that can’t really buy us love and can’t really buy us happiness. And for a long time the research of the scientific evidence that we had was actually confirming that assumption. Because we know that GDP in countries like UK went up over the last decades but the happiness just simply did not and the life satisfaction did not, and the same was true on the level of the individuals.

So above a certain threshold that was very low, increased income or increased spending didn’t really do anything for people’s happiness. So for a long time the conclusion was really that money can’t really buy us happiness. However money has actually recently come back to the table because researchers suggested that maybe if money does not buy us happiness it is because we are not spending it right.

And they showed actually that if money can buy us happiness if we spend it on experiences rather than material goods or on others rather than ourselves. And now coming to my own research what I have suggested is that basically spending it right means spending it in a way that is in line with our psychological needs and preferences. So what we shown is that if people allocate their spending in a way that matches their personality they are actually happier.

So you can think of an extrovert spending their money on going to the pub having a good time with their friends, experiencing something stimulating and exciting. And an introvert rather allocating their spending in a way that helps them make the most of their quiet me time that they want to have, such as reading a book.

02M28 Personality centric spending, targeting and visualizing are main fields of your research interests. What made you decide to go into them?

Yes, I think what I find most interesting about this line of research is really that consumption is something that affects all of us. So whether we like it or not it is quite a big part of our everyday life’s. As a psychologist what is really interesting to me is how it affects us and how it effects our well-being.

And what we can do to make the money work for us rather than constantly worrying about working towards more money. The whole field of personalization and customer-centric marketing which is very closely related to that, is so attractive to me because it is so fast paced and so cutting edge.

So pretty much every day there is new technology coming to the market that’s aimed at helping us find the stuff that we are really interested in. I think as a psychologist we really have the opportunity to contribute to that and to find ways of actually bringing back this human element to personalization and the digital world that helps us to make better decisions and really find the stuff that is interesting to us.f

03M41 You did some research on spending money and well-being. If you look at it, what can be applications of that research and what parties might benefit?

I think, one of the main application of my research is really customer-centric marketing.

And what we have seen over the past few years is a continuous shift from what is called product-centric marketing. That is companies start with a product and they try to optimize their marketing strategy around the product.

And it is basically this continuous shift towards a more customer centric approach, where businesses really start with the customer they try to optimize their services by focusing on the individual needs and preferences. What is really crucial in that context is that you understand these needs you understand what is it that makes customers happy what it that doesn’t.

And in that sense the idea of customer-centric marketing is really one of dual value creation that creates value for companies but also for their customers. So on the one hand yes it helps customers to find products and services that are of interest to them and that also help them to lead happier lives.

But of course companies are not so like the ultimate goal for the company is probably not make the customer happy but to increase their profits. And again the idea here is that if you have a happy customer it’s the one that will create the biggest value and for company in the long term because they turn into loyal customers and they do not just purchase once but they come back again and again and again.

05M12 Your research is predominantly done online and moreover in the social media. What does this mean for the application of doing research offline, in the ‘’real’’ world so to speak?

Well, I think that actually the fact that we are doing research online gives us a chance to capture the real world much better, than what was previously possible offline. If you take psychological studies and the majority of them has always been done with under graduated students in the lab.

And of course that is great in the lab we have control over the conditions that we run our experience in and the also we can test for something like causality. But then on the other hand it also means that it often times it is quite far from the ‘real world’ and the findings that we get in the lab might not necessarily generalize to the real world and actual people.

So the fact that we can now actually study people online gives us the chance to see what it is that people are actually doing in their day to day lives. And to capture these real behaviours and what is interesting about this, is that often times the stuff we observer online is just a mere expression of what we experience offline.

So if we see that someone is listening to Adele on Spotify or has checked in at Starbucks on Oxford Street this is a digital trace but it actually reflects the actual behaviour that happening offline.

06M39 Why is it that personality centric targeting results in these significant increases in CTR (click through rates) but moreover gigantic surge in conversion rates?

Yes, so actually when I looked at the findings for the first time I was really surprised to see that the effect of something like psychological targeting was bigger on conversion rates then it was through click through rates because if you think about the customer journey, they basically they click first and then if anything they purchase. So I would have expected that the effect to be bigger on click through rates I thought about it quite a lot of why we find these patterns.

And I think the reason is probably that first of all clicks are really noisy especially when I am using my phone, I often accidentally click on something something that I didn’t want to click on, so there is quite a lot of error around these clicks. And second of all clicks are also something that are very spontaneous and it is just the first expression of an interest.

So I might think yes that I am interested in playing a party app after having seen the ad maybe I like the picture. But then once I get to the google play store and read more about the app and the description, I probably realize that it is not for me because I am an introverted geek and I do not even like partying or drinking.

So in that sense conversions are in comparison to clicks are much more effected by like this deliberate choice or deliberate decision to actually download something to buy a product. And of course at the moment this is just an educated guess and I think we need to go back to the lab and test that empirically and see what is really going on here.

08M14 What is the difference according to you between consumer centric on the one hand and the personality centric on the other hand in terms of approach? And what do companies need to have in place in order to make this move from customer to personality centricness?

So I think there isn’t actually really a difference between the two. Customer centricity is really this broad frame work that describes how the customer experience can be personalized can be made more appealing. And personality is one variable that can help us actually achieve this personalization.

So at the moment companies are mainly relying more on something like demographic variables, age, gender or past behaviors, something like a past purchases that a person has made or the websites that they have visited.

But I think the more and more progress we make like predicting the psychological traits, the more and more personality variables or other psychological variables will become important in supporting this customer centric approach.

09M12 In some of your research we see that you target on one big 5 factor only. In order to improve the effectiveness of personality centric targeting, how many of the big 5 traits should at least be targeted simultaneously?

Well I think it is hard to say and at the moment we do not have any empirical evidence for that question but I do not think that there is a minimum or even an ideal number for traits that you should target simultaneously, because at the end it depends on the product.

So if the product has distinct characteristics on one, two or five traits and this is something that we should take into consideration. And I think it might very well turn out that of the traits are more effective. So we know for example extraversion has traditionally been considered as one of our traits that is most important in explaining and predicting behaviour.

Similarly it is also dependent on the situation, so for example we might find that openness works particularly valid well in the context of music preferences and movie preferences. While a trait like conscientiousness can actually be better at predicting insurance needs or insurance preferences. So in the end I think there is no generic answer and I think it really depends on that specific product and specific situation that you are dealing with.

10M34 Let’s assume I were a marketer and I would start with personality centric marketing and I had a limited budget. What would you advise me to spend my money on: On personality centric messages or personality centric product fit?

So again I think in a way it depends on the product that you are selling so from the research that we have done so far it seems that product targeting is more effective in terms of increasing conversion rates. But that only works if you have a product that has distinct characteristics.

So if you sell gardening equipment for example then it is quite likely that you might get the biggest effects if you target people who are introverted and who are conscientious. But the thing is that a lot of the products simply do not have these distinct characteristics.

Just take a mobile phone for example. In a way everybody uses mobile phones and they are kind of very generic in terms of their functionality and what they can do. So unless there is a strong brand personality associated with that phone that kind of artificially creates a personality profile, you might be better off in simply doing what I call message tailoring.

So you can start with any audience you like. So you can say, ok I want to sell my phone to extroverts today. How can I sell this phone to this type of person, to make it as exiting and appealing as you possibly can? But then in the end of course ideally you would combine the two because it is not even that much more effort to say, ok I have this product, that has distinct characteristics, I know what my target group looks like and now I simply try to create my advertising material in a way that corresponds to that target group.

12M16 Already in 2012 your fellow researchers Kosinski and Stillwell showed that personality centric targeting was a next step in marketing. And now your latest research seems to be indicating that the benefits are significant in fact huge. How come, do you think that industries like retail hardly seem to make use of these findings?

Yeah I think one of the reasons that psychological traits haven’t really been used in customer centric marketing so far, is that they are latent constructs. So there is nothing that we can measure and observe directly. And for companies it is pretty easy to collect data on someone’s demographic variables like age and gender, or just collect data about what they did in the past or what they have purchased in the past or what websites they have visited.

But it is much harder to get actually to something like psychological traits. So yep, my colleagues Michal Kosinski and David Stillwell showed that it is possible to predict personality from data that actually should be available to companies already. But the thing is that you still need to have the models or access to the models and algorithms that turn those data into psychological predictions and that’s something that companies simply don’t have at the moment.

But again here, there are so many new tech companies entering the market and providing these commercial predictions of personality from really any kind of data that you can possibly imagine. So I think that basically companies will get access to these models and as soon as they have that, psychology or psychological traits will become much more important in the context of customer centric marketing.

14M38 Early December 2016 the Swiss magazine ‘Das Magazine’ sparked discussions about personality centric targeting and election results, the American result even. What can be the consequences of this discussions for the application of your research findings and the research itself?

I think a lot of the discussion that originated in the article has really been around the question of whether or not Cambridge Analytica and their psychological targeting helped Trump to win the election. And my opinion on that is that it is pretty much impossible to know because we don’t have any access to the campaign details and we also don’t have data on the campaign outcomes.

And the fact that we simply cannot answer the question means that in a way the whole discussion that evolves around the article, in my opinion is somehow misguided. Because the point is that these technologies exist and that we can now implement them at an extremely large scale as the Trump campaign shows.

We can implement them without the public even being aware of their existence, so I think the discussion that we should be having instead is really one of how can we regulate these technologies to make sure that they are being used to the benefit of the society rather than against society.

I really hope that the article and the publicity that it got, really helps us to get the public attention and the public support. That we need to convince policy makers that it is time to act now, and that it is really time to come up with policies and regulations to make sure that we control these technologies and use it in the interest of society.

16M23 I will try to wrap it up, your fields of interest are personality and well-being, personality and advertising effectiveness and personality and advertising aesthetics. What are the three important and unanswered questions in these fields and what does that means for your research priorities?

Well, there is so much I actually want to look at but if I had to boil it down to three questions. I think the first one would actually be to see how effective personality or psychological targeting is in comparison is to existing approaches. So we know from the results that we have so far that it is effective, but we do not really know how effective it is in comparison to demographical approaches or to behavioural approaches.

And also how effective it could be if were to combine these approaches. And the second question that is really interesting to me is to see if we can learn even more about someone’s motivations and needs if we take the context into consideration. So if we know that an extroverted person is in an extrovert situation does it make him more or less likely to respond to a personality targeted add.

And what is so interesting here is that we actually now have the opportunity do that because all of us carry around smart phones and these smart phones are just amazing technologies to see what is happening around us.

So we have microphone data that tells us about how much conversation there is going on or we have the accelerometer data that tells us how much physical activity there is. So in a way we already have access to the data and now the question is how can we make sense of the data that we get and how can we use it in combination with more stable psychological traits.

And then the last question which I think is probably one of the most important ones is, to look at the public opinion towards something like psychological targeting. Because in the end it is supposed to help consumer to find the stuff that they are really interested in.

But obviously there is also the danger and we have seen that in the context of the Trump campaign that the technologies are being used without the people even being aware of it. So what I am interested in is to find out a way to maintain people’s trust and to implement these technologies in a way that people accept them and consider them a useful service rather than like an invasion of their privacy.

Interview Peter van der Bel, founder & curator of The Centre for Applied Product Personality Research