776841

research-article2018
MRE0010.1177/1470785318776841International Journal of Market ResearchWirth

Viewpoint

International Journal of

Hello marketing, what can artificial

Market Research
2018, Vol. 60(5) 435­–438
© The Author(s) 2018
intelligence help you with? Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/1470785318776841
https://doi.org/10.1177/1470785318776841
journals.sagepub.com/home/mre

Norbert Wirth
PricewaterhouseCoopers GmbH, Germany

Abstract
Artificial intelligence based applications are emerging in a broad range of expert domains. News
about AI based solutions in medicine, industrial production processes, logistics, mobility and
digital marketing trigger discussions and a lot of speculation. The market research industry seems
hesitant and at the same time eager to embrace this new technology. In this article the author
provides a definition of artificial intelligence and its different forms: narrow AI, hybrid AI and
strong AI. He concludes his reflection on the question whether it’s feasible to develop AI based
marketing insights solutions with the recommendation: it’s time to embrace AI.

Keywords
Artificial Intelligence, AI, Marketing Insights, Market Research, Narrow AI, Hybrid AI

Over recent decades, we have become accustomed to hearing or reading about dramatic predictions
on how technological developments will destroy, or at the very least dramatically transform market
research. If not social media, then definitely big data would undoubtedly mean the end of market
research as we know it; with its roots in scientific methods and empirical research.
Artificial intelligence (AI) is the latest big game changer. The number of AI start-ups is growing
rapidly, there are more and more conferences dedicated to the topic, and we hear a lot about big
tech companies investing huge amounts of money in AI-related developments. Somehow, this does
not come as a surprise, as there has been a lot of buzz about machine learning, deep learning, and
analytics over the last couple of years—and that is what AI is about, right? All these developments
are alluding to the simple fact that data alone, hailed as the “new oil,” is not doing the trick.
However, just labeling a solution, AI won’t make it relevant or successful. What does AI really
mean? Is it just a synonym for extremely powerful algorithms and neural networks or is there more
to it? The current hype around AI is creating a blurry picture calling for clarification. Complex data
landscapes can be explored to understand what information they provide. However, data will only
translate into significant business potential if related to the right use case. Similarly, the use case
determines the appropriate analytical approach and degree of automation.

Corresponding author:
Norbert Wirth, PricewaterhouseCoopers GmbH, Friedrichstraße 14, 70174 Stuttgart, Germany.
Email: [email protected]
436 International Journal of Market Research 60(5)

I’ll refrain from creating a dystopic vision of the future, and instead focus on the question
whether AI—in in its current development stage—can replace or augment the expertise and meth-
ods it takes to inform marketing decisions. The latter is, at least in my humble opinion, the essence
of market research. And if AI is powerful enough to inform decisions such as will person A like
product B and will consumer X purchase the car Y at price Z, then it’s overdue to embrace it.
Especially if it’s doing a better job than human experts. The remaining step for such a system
would be to automatically trigger the best possible marketing action and continuously learn from
the market response.

Back to basics
To understand what AI is capable of, it makes sense to look at the basic definition first:

Artificial intelligence (AI) is intelligence exhibited by machines. In computer science, the field of AI
research defines itself as the study of “intelligent agents”: any device that perceives its environment and
takes actions that maximize its chance of success at some goal. (Wikipedia, 2018)

There’s a lot of information in these few lines. The first part can be translated into: “[. . .] com-
puters do things that require intelligence when done by humans” (Copeland, 2000). So, it’s easy
to crack the “artificial” in AI; this obviously means machines/computers are doing something,
not humans. The tricky bit is the intelligence. In the very early days of AI, before John McCarthy
held the famous “Dartmouth summer research project on artificial intelligence” in 1956, Alan
Turing (1950) wrote a wonderful paper under the visionary title “Computing Machinery and
Intelligence.” The first two sentences of Touring’s paper are remarkable—and they give us a
solid clue:

I propose to consider the question, “Can machines think?” This should begin with definitions of the
meaning of the terms “machine” and “think.” (Turing, 1950)

Intelligence?
What constitutes intelligence or in other words, how can a machine possibly think? Human beings
think all the time, but it’s not easy to distinguish thinking from purely mechanical processing.
There’s no point in drifting into a philosophical discussion on this; let’s focus on the required ingre-
dients of the “think” component; representing the “intelligence” in AI.
The following are more or less application independent:

•• Learning;
•• Knowledge representation;
•• Reasoning;
•• Prediction/planning.

Depending on the specific problem at hand, we’ll need a lot more; for example, the particularly
tricky emotional intelligence. Back in 1950, each of these ingredients must have been considered
a major challenge. However, today, in the age of Apple’s Siri and Amazon’s Alexa, AI has obvi-
ously evolved and many problems have been solved. In fact, in today’s terminology, we distinguish
different evolutionary stages of AI, ranging from application-specific, narrow or weak AI to
Artificial General Intelligence (AGI) or strong AI.
Wirth 437

Today’s perception of AI in the general public is massively influenced by movies like Her or
Ex-Machina. “Super” intelligent systems easily outperform human intelligence on pretty much all
fronts and speak to their inferior human contacts (the word “user” seems inappropriate in this
context) with the voice of Scarlett Johansson—the latter at least in Her. This type of AI is referred
to as AGI, or even Superintelligence, and it’s fair to assume that for the time being, it’s science
fiction.

Narrow AI
Weak or narrow AI is tailored to a specific problem or task and cannot deal with other challenges
without being re-trained and/or modified. Calling this weak AI is a little misleading, as it’s by no
means easy to develop such a system. The name only makes sense if we look at the following ques-
tion: How strong and flexible is the system in relation to human intelligence? Weak or narrow AI
systems lack the flexibility of human intelligence; they fall short in terms of the scope of compo-
nents that comprise human intelligence but can be very powerful in their domain. In fact, they typi-
cally aim to beat humans in their specific domain, and once they achieve this, they get a lot of
media attention. DeepMind’s (now Google/Alphabet) AlphaGo, or IBM’s DeepBlue are popular
examples. Pretty much all AI that is operational today falls into the narrow AI category: Apple’s
Siri, which goes back to a SRI International spin-off acquired by Apple and a speech recognition
engine developed by Nuance Communications, Google’s Assistant, and Amazon’s Alexa. There
are plenty of narrow AI solutions addressing use cases in numerous industries ranging from health
care to defense and—spoiler alert—marketing!

Strong AI
Strong AI, full AI, and AGI are being used synonymously. A Strong AI system is as powerful and
flexible as human intelligence and is not tailored to a specific problem or task. Strong AI has not
been achieved (yet).

Hybrid AI
Recent developments are calling for a more granular terminology than the two extremes of narrow
AI and full AI. We can observe rapid growth of narrow AI with applications for an increasing range
of use cases and on top of that the emergence new extremely powerful AI systems; some of them
blending multiple narrow AI solutions, capable of adapting to new challenges. The latter are still
not strong AI; however, more than “just” narrow AI. Call it semantics but I simply wouldn’t want
to put IBM’s Watson where it stands today or the AI engines of companies like Google into the
same bucket with a chess computer on steroids. AI solutions that combine multiple narrow AI
modules are referred to as hybrid AI (Greenwald, 2011), and this field is growing at a fast pace
(Martínez de Pisón, Urraca, Quintián, & Corchado, 2017).

AI solutions for marketing

Marketing and Data Scientists can leverage a very rich toolbox of predictive models, machine
learning approaches, and the like. I wouldn’t refer to these as AI, although some of them can be
leveraged when developing AI solutions. We’d expect an AI solution to be capable of learning,
knowledge representation, reasoning, and prediction or planning. All of that can be focused on a
438 International Journal of Market Research 60(5)

narrow field like language, or image processing. The solution can naturally leverage multiple nar-
row AI modules—then we’re talking about a hybrid solution.
The availability of ready-made AI components, such as language processing or image recogni-
tion, that we can access free of charge or for a license fee is amazing. More services and application
programming interfaces (APIs) are emerging constantly. What’s even more impressive is the open
access to toolboxes like TensorFlow, with very user friendly interfaces via common analytical
platforms or languages such as Python or R. This enables Data Scientists to leverage very powerful
tools and algorithms without even leaving their comfort zone. They don’t need costly on-premise
technology stacks because the well-known cloud services can be scaled up as needed. So, if some-
one has a clearly defined use-case and knows how to use what’s available, it’s not rocket science
to create a hybrid AI solution.
This leaves us with the question if an AI solution can replace the human expertise required to
generate valuable marketing insights, consequently trigger respective actions, and continuously
learn from what it’s doing. Well, of course this is already happening in areas such as online target-
ing and dynamic attribution, next best offer solutions—the list goes on. No surprise that we see
these solutions growing rapidly in the digital ecosystem because these decisions are made at such
scale and speed that a human would be simply lost. And it doesn’t stop there. Training an AI system
so it learns the association between brand preference and consumer profiles is just another use case
and adding a nice voice interface is the icing on the cake. So, “hello Marketing, what can AI help
you with?” It’s time to embrace AI.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.

References
Copeland, B. J. (2000). What is artificial intelligence? Retrieved from http://www.alanturing.net/turing_
archive/pages/reference%20articles/what%20is%20ai.html
Greenwald, T. (2011). How smart machines like iPhone 4S are quietly changing your industry. Retrieved
from https://www.forbes.com/sites/tedgreenwald/2011/10/13/how-smart-machines-like-iphone-4s-are-
quietly-changing-your-industry/#46547361598f
Martínez de Pisón, F. J., Urraca, R., Quintián, H., & Corchado, E. (Eds.). (2017). Hybrid artificial intelligent
systems: Proceedings of the 12th international conference, HAIS 2017, La Rioja, Spain, June 21–23, 2017.
Cham, Switzerland: Springer.
Turing, A. M. (1950). Computing machinery and intelligence. Mind, 49, 433–460. Retrieved from https://
www.csee.umbc.edu/courses/471/papers/turing.pdf
Wikipedia (2018). 14 January. Retrieved from https://en.wikipedia.org/wiki/Artificial_intelligence