The marketing AI revolution has arrived… almost
Editorial

The marketing AI revolution has arrived… almost

Artificial intelligence (AI) has suddenly become a mainstream concept for marketers. Although they’re not quite sure how they are going to use it yet, marketers hear about this topic again and again

But as a course director at the Chartered Institute of Marketing, I know the kind of information that marketers - time poor and hungry for details - truly need.

So, let me talk you through some essential ways in which marketers can already use AI, and some of the technology that is just on the horizon. We’ll also explore the potential pitfalls of AI in marketing and how to avoid them.

Where is the excitement around AI coming from?

A lot of the buzz around AI in marketing comes from a wave of announcements released by Google this year on how they will be using AI, or Machine Learning (ML) as they tend to refer to it in this context, to improve their ad products.

Google make over 90% of their money selling advertising. Since most of that ad revenue is paid for on a Cost Per Click basis (that is, they only get paid if someone clicks on the ad), they are very keen to only show you relevant ads so that you make that pivotal click.

Rather than purely rely on advertisers getting their targeting right, they are now using ML to serve the right ads to the right people. As well as knowing what phrase you’ve searched for, they can now also factor in things such as your previous searches to try and understand the context of why you are searching.

This means that marketers using Google’s ad products may be using AI without even knowing it. It also means that AI technologies are being widely discussed and, as a result, industry awareness is growing.

Has AI been delivering for marketers?

A recent and fairly short-lived buzz phrase in digital marketing was ‘Messenger Bots’. These bots were claimed to be able to provide customer support and product discovery via chat apps like Facebook Messenger. This was touted as the latest form of AI in many quarters, and was hotly talked about at conferences and online.

However, most of these messenger bots just offered replies to users based on scripts (i.e. If you ask this, say this, etc). Clearly this is very far from true AI, and often leads to a disappointing user experience.

This, and many other false starts, has led to a certain level of cynicism about AI in the industry, and only furthered the confusion as to what it really is.

However, it seems that there may be positive news on the horizon: Salesforce’s annual state of the nation report shows most marketers are investigating and trying to use AI more widely, while 57% of marketers currently using AI say it has become essential to their strategy.

Practical uses of AI in marketing

In an incredibly noisy online environment, anything that gets the right content at the right time in front of the right audience is being taken very seriously.

Whether it’s using ML for recommending the relevant products on ecommerce sites or using AI to refine email spam filters, it’s all about improving the user journey online.

One of the most high-impact practical uses, and one of the few examples of a true AI that audiences can play with right now, is the IBM Watson Twitter analysis tool. By plugging in your Twitter account, the IBM AI will analyse your tone of voice and the character of your brand. This is incredibly useful for working out if your social activity is on-brand and how you might improve things.

Practical examples such as this are having a big impact and demonstrate the insights AI will bring to marketing in the near future. But, as with any technology, while the opportunities are huge, in the early stages at least, the risks are very high too.

Microsoft’s AI Twitter account, the ‘Tay’ chatbot, launched to much fanfare but was quickly switched off after it became sexist, racist and genocidal within days of launching. Those of us who understand how AI is trained could have seen this coming, but it creates concerns for those that don’t.

What we need is a broader understanding of the underlying technologies and techniques, and more practical tools that can be used by non-experts. It’s a fascinating and exciting time for AI in the marketing world and we’re on the cusp of mass adoption. But there are a few steps still to take…

Discover more from Daniel and other industry experts in the latest Access-AI e-book: AI & Marketing.

Daniel Rowles Course Director CIM
Back to all