Artificial Intelligence in Marketing
November 28, 2018
Get used to AI marketing
Long gone are the days when you used to load up a clothes website and have to type precisely what you were looking for only to get one result. “Black men’s coat” may well just return the only coat that was described specifically as black in the title. Now when searching through the big fashion retailers it will show you the most relevant results to your search term as well as any other relevant items that may match it or that are regularly bought with it and allow you to filter down your results depending on price, size, brand or any other number of parameters. Pointing this out seems a little rudimentary given that it’s become the norm and exactly what you’d expect but those of us who remember the pre-Facebook internet can cast our minds back to a time when there was very little AI presence of any kind.
Intuitive search results are a very basic demonstration of marketing AI but give you perhaps a relatable example of just how engrained into marketing AI actually is. As we move towards an age where some predict that AI will dominate most aspects of our lives so too must we accept that it will begin to dictate much of what we do in marketing and prepare for the benefits.
Whether via Facebook marketing tools that allow us to predict consumer behaviour or to narrow our campaigns to impossibly specific margins, we use AI in most aspects of our jobs already, we probably just forget that we do. Perhaps it’s easier to pretend that we don’t.
As a consumer your data and behaviour is your most valuable asset. Back in the early days marketers would rely heavily on consumer participation in questionnaires and focus groups in order to gain market insight. The biggest problem with this type of research is that it isn’t natural and people tend to either exaggerate or flat out lie about their true intentions in an attempt to mask characteristics they deem undesirable such as selfishness.
Where these types of research failed AI excels. With the right software you’re now seeing, in real time, exactly how much time consumers look at your products, where they look, how far down the page they’re willing to go and any number of other data fields you’d care to include. We’ve now reached the point where there isn’t really any difficulty gaining data on consumers it’s just about having the right instinct to know what to use and how to execute it.
Marketing AI infiltration
Customer service is a fine example of modern business employing masses of data collection to its advantage to predict customer behaviour. Most of the large entertainment companies now use either chat or messaging software to speak to its customers. Such software is able to analyse the language, frequency and subjects that customers discuss and predict with 88% accuracy the next time they’ll get back in touch and which products they’re likely to take out.
The UK Press Association has partnered, for example, with a project called Reporters and Data and Robots (RADAR) to have robots write over 30,000 local news stories after being fed with government and local authority data. They’re programmed to use natural language and writing styles and have so far been an enormous success. At the risk of writing myself out of a job, it’s not difficult to see content writing robots being fed masses of data in order write the most mathematically engaging content available. Having said that, I’d argue robustly that we’d miss the human element of interaction via writing and language.
In many ways it’s an uncomfortable truth that not only will our lives likely be dictated to largely by robots and AI in the future but that we’re almost in that situation anyway. If politically malevolent forces can predict your sexuality with 91% accuracy already then it’s tempting to just give up and accept the inevitable. It’s not all doom and gloom though and we should, as innovative professionals, embrace new ways of working and serving our clients in the best way possible.