Artificial Intelligence (AI) has hit the mainstream with undeniable force over the last year, with practical examples of its application beginning to manifest themselves within the business world, and industry analysts such as Gartner quick to lead the conversation on its vanities and virtues.
Within the marketing world, that has fuelled fears AI will take people's jobs, and there is even talk of taking action to prevent AI from impersonating humans. As with every marketing craze, however, before whipping out a chequebook, brands need to take a step back and really understand what these solutions are capable of, and what the limitations are.
For starters, it's important to realise that there is no single, all-encompassing "AI" – it's a term that refers to a field of research and an approach to computational and algorithmic design. As each "AI solution" needs to be designed to fit the problem at hand, each application will be unique. This not only means that companies need to invest in training their AI to solve for specific problems (i.e. vertical AI solutions), but also that failing to do so and relying on generic AI (horizontal AI) for every problem can result in misleading and even broken results.
Getting to emotion with AI
A key area that marketers will need to tread cautiously with when using AI is arguably one of the most important parts of branding: emotional tone and language. This area can really make or break the success of a brand; the emotional language in marketing messages accounts for as much as 60% of response across channels – email, Facebook and display. Put simply, emotions in language are generally more important than price points, format and message structure.
So how does AI fit in with this?
In one practical use case, IBM Watson can be used to detect emotion in a telephone call. For example, if an agent was told "somebody created an account using my email account," Watson would detect Anger (0.76%) as the main emotion, and a lower 51% if somebody said "this is not my account".
This would arm the agent with emotional cues for use both on the call or for follow ups. This intelligence equips agents to deal with the situation more effectively and provide a more personal customer experience.
Taming the beast
However, when it is not trained towards a specific goal, AI struggles to be effective at all, let alone powerful.
This is an extremely important point for marketers to realise, as the main issue with current models is that they are not specifically created for, or adapted to, marketing. If you turn to a generic AI like IBM Watson's horizontal API and apply it to marketing messages, you'll get little to no actionable insights; in fact, you may even get the wrong insights. For example, the programme says that the message "Attention please. Our offer ends today" is a rather "sad" (77%).
Does it sound sad to you? Of course not. In the context of your mailbox, this subject line's emotional motivators are actually Anxiety (attention please) and Urgency (offer ends today). Watson is limited to five generic human emotions and tries to understand what emotions are conveyed, but this is too simplistic to work within the marketing context.
To solve this particular problem, marketers need an AI that has been trained for the specific purpose of figuring out what language and emotions inspire a given person. This may sound futuristic, but it is possible today. Such an AI could analyse customer responses to communications, assign an emotional ID for each customer, and then suggest the suitable language that better resonates with each of them.
Training complete, let's talk
It's a bit like when you're having a conversation with someone and adjust your language based on cues you detect from them. AI can recreate that capability for digital channels so that you can communicate with millions of people at once and remember what works for each and one of them.
One-to-one personally relevant communication is the holy grail of marketing. When correctly associating AI language and emotions to drive personalisation, the results can transform the way companies and their products interact with its prospects and customers – both in terms of experience and business results.