AI marketing to extroverts — The rise of machine marketing

I don’t usually refer to serious scientific research as hilarious, but predicting the personality of extroverts opens a whole sector of issues. Using AI to do so makes it funnier, but there are so many human ironies.

Research by Nanyang Technological University, Singapore (NTU Singapore) is exploring personality traits in marketing using AI to predict consumer wants based on their online behaviour. An early finding is that extroverts may be “attracted to marketing messages which match their personality”. This is marketing linguistics, and it’s a tough range of subjects.

This is my area of expertise. I’m a commercial writer. I have 11 million words online paid for by other people. Everything I do in commercial writing is about messaging. So when I see this sort of thing, I do have pretty strong opinions.

My first reaction to this was “Ah, machines marketing to vegetables! Huzzah! Throw another trilobite on the barbie!” and to go and get some coffee. I’m not the world’s leading idealist in this field. The sheer volume of verbose useless garbage in content marketing is beyond obscene.

To be fair, though, this study is a lot more complex. It’s a huge field. Prediction isn’t as straightforward as drivelling psych hacks think it is. Machine learning does mean number crunching, but you do have to do more than just crunch numbers.

To get anywhere at all in this area of research, which is standard market research in some ways:

• You need to make the terminology and the numbers functional.

• You need to prove they’re functional to the people paying for them.

READ  How to Make a Paper Plane with Science - Tech News

• The metrics can be murderously detailed, per word or expression, and despite urban myths, all aspects of marketing can be measured very accurately.

It’s not easy. It’s pretty thankless work. It’s unspeakably time-consuming. That said, this is an interesting take on the usual mindless and endless buzzwords approach.

There’s another obstacle here which is much less obvious. It’s marketing methodologies. Content writing and marketing is often mindless. “Brand voicing” is a case in point. Voicing can and often does fossilize into some ancient relics and curios show. That’s why you see so much rap and “Yo! Dude!” and other 1980s rubbish still lumbering around. This stuff is 40 years old and much other marketing content is a lot older. It’s meaningless gibberish or accepted babble to people under 40, and those over 40 don’t give a damn about it. It’s not information, it’s a sort of content-based insult.

Extroverts as marketing subjects

Marketing to extroverts may seem redundant. Aren’t these the people/morons who rush out and buy the latest crap at the highest prices and yell at people how great it is? The people whose ancestries are based on buying every fad?

Well, yes and no…

Extroverts are a bit more complex than they seem, as though there was a choice. Extroverts are rewards-based people. Their dopamine systems get a charge out of rewards, and they respond positively to mixing with a lot of people. The brains of introverts and extroverts are fundamentally different in that way.

As marketing subjects, however, extroverts are besieged by choices of rewards. That’s what this NTU Singapore study is all about. There’s a lot of competition for their attention. The study is right on the money, pun intended, when it comes to terminology to connect with this market.

READ  The physician whose 1964 vaccine beat back rubella is working to defeat the new coronavirus | Science

Consider these two lines:

• It’s chic! It’s elegant! It’s your fabulous new car!

• These wheels bang big! It’s the ride of your life!

“Chic” and “elegant” in the first line may mean nothing to an extrovert. Some extroverts do have vocabularies, but can’t often find a use for them. The second line may refer to sex, which some of them have apparently heard of, although it’s unclear how they find time to reproduce. So the second line would be the natural choice to market to them.

The NTU Singapore research

OK, now try proving that to a corporate client with a multi-million dollar marketing budget. The idea of the NTU Singapore research is to develop a predictive ability for marketing purposes. The linguistic approach is 100% right. Terminology has to be efficient, and connect properly to the audience.

It’s basic communications 101, as well as marketing 101. You talk to people in the language that connects with them. The research has found that “positive emotion words” (environmental terminology like “good”, “beautiful”, etc.) and “social process words” like meet, talk, etc. and other socialized verbs like “we” perhaps have some connection here. Pronouns except “I” for some not-very-well-defined reason, show social intentions, also well in the extrovert spectrum.

If you’re in the sector and you’re thinking NTU Singapore has opened a can of something with anything resembling technical, let alone psychological terminology, you’re right. Most marketing people do have some psych knowledge and training. Their C level people may have some, too.

However – Try translating machine learning and AI into this field, and you have an unavoidable encyclopedia of terminology to work with. Predictive algorithms are big money business these days. This can be a very difficult sell, and you’re selling to people who do know how to say No when it suits them.

My advice to NTU Singapore would be to translate into hard numbers which make sense on a spreadsheet. It’s not just the extroverts who need to be marketed to; the C level guys need their own predictive algorithms to see which way they’ll jump.

The most beneficial effect of this research, ironically, could be in baseline communications. How do you get a message through in good working order to people who need to see it? With terminology. That’s the key. This is the early pre-Cambrian of machine marketing, and the steps are many. Be patient, NTU Singapore, and you may just revolutionize marketing communications.

This opinion article was written by an independent writer. The opinions and views expressed herein are those of the author and are not necessarily intended to reflect those of


Leave a Reply

This website uses cookies. By continuing to use this site, you accept our use of cookies.