Some thoughts on Artificial Intelligence, and what it means for marketing

I am preparing for a workshop on the topic of Artificial Intelligence in Marketing, and what it means for the field. These are some initial thoughts that I penned for this workshop.

What is Artificial Intelligence (AI)?

The term AI refers to any technological assemblage that can collect inputs from the environment (e.g., through sensors), and take actions as a result of those inputs (e.g., adjust temperature), in ways that simulate human intelligence. This means that the technology can apply rules, can self-correct, and can learn through the acquisition of new information.

 

This definition means that AI is not about a particular device or even a set of devices. But, rather, about what the combination of devices can achieve together. I.e., a machine learning algorithm, per se, is not AI.

 

It also means that there isn’t an objective measure for what is AI or isn’t. If current AI technology had been tested in, say 1960s, would it have been able to pass Turing’s test?

 

What are some examples of applications of AI in marketing, and how do they impact on a) consumers and consumption and/or b) the practice of marketing?

As per Huang and Rust (2018), AI can be developed for a pre-defined context and task – e.g., for the purpose of product recommendation – in which case they are called weak or narrow AI. Alternatively, AI can be developed for broad application, meaning that they are expected to be able to handle (i.e., make decisions / take actions / reason about) unfamiliar situations. Such applications are called strong or generalised AI.

 

The Turing test, sci-fi movies, etc… tend to focus on strong AI, which makes AI seem like a scary but distant reality. Weak AI is, actually, quite common in our lives, though most people may not think of those applications as AI.

 

Application	Impact on consumers and consumption	Impact on marketing management
Recommendation systems – e.g., Amazon, Spotify, Youtube…	Customer is presented with products that complement or substitute past purchases or searches.   Could speed up buying process.   Could encourage additional spending.   Could also reduce choice / reduce serendipity / create bubble.	Could encourage cross / up selling.   Could help with stock management.   Most effective for companies with large portfolios?
Customer service bots – e.g., Amazon	Could improve access for some groups – e.g., deaf, social anxieties.   Could hinder access for some groups – e.g., blind.	Screen customer support contacts to most qualified person.   Cost savings.   Challenges of spelling mistakes (if written) or accents (if audio).
Advice robots – e.g., law, mental health…	Make services accessible to new segments due to low costs or remote accessibility.	Could support stretched staff (e.g., NHS)
Emphatic AI for ad display – e.g., Jaguar / #FeelWimbledon campaign	Surveillance   Privacy   Manipulation	Support brand attitude, by avoiding negative associations
Smart devices – e.g., Car	Improved safety.   Consumption of entertainment / news.	Easier maintenance.   Insight from usage data.   Issues of data security.   Could encourage cross / up selling.

 

What are the most urgent and/or important research questions that marketing scholars should tackle?

Some urgent research questions that are particularly appealing to me, given my interest in customer profiling and targeting are:

• Which customer segments are excluded / discriminated against / underestimated due to biases in automated decisions (due to biased datasets or algorithms? This is particularly important for applications that scale rapidly.
• What is the impact on role of marketing and marketers in the organisation (e.g., customer insight)?
• What is the impact of the behavioural and managerial changes discussed in the previous point, in terms of the skills needed for those in a marketing function?
• What is the impact of these changes in the organisation’s ability to innovate?

 

What is / are the key challenge(s) faced by marketing scholars in pursuing the research questions identified?

In my view, the key challenge relates to the opacity of AI, not just in terms of how decisions are made but even in terms of awareness of when AI is being used (e.g., Google Duplex).

 

This opacity means that we can study what is happening, but will struggle to study how that is happening, and even more to understand why that is happening. While marketing academics have always played catch up when it comes to understanding technology (e.g., regarding digital technology), AI is at a whole new level of challenge in terms of access and complexity.

 

What are your thoughts about AI, its use in marketing, and the implications for marketing scholarship and practice?