Have you ever wondered how it is that you look at something on Amazon that then, suddenly, that item pops up in ads on every other webpage you visit? Or why it is that there are always adverts popping up on your Facebook and Instagram feeds that relate to something you’ve bought before? You didn’t imagine it; there is a connection, Artificial Intelligence (AI).
The strides that AI has taken over the last few years have been awesomely impressive, and some of the most successful digital companies have been able to create and implement systems that learn customer behaviours to personalise product and content recommendations. In most cases it works, Netflix has mastered the ‘you liked this show, why not try this’ method of keeping us binge-watching programmes about serial killers, and Tesco and Sainsbury’s have managed to work out that if you buy cat food, you may also need pet insurance.
This hasn’t stopped some well-publicised faux pas, though, such as replacing online shopping orders for daffodils with spring onions or popcorn with potatoes, which brings up the question: should we entirely rely on AI, or do we still need to use some human intelligence?
Don’t get me wrong, there’s nothing wrong with brands trying a ‘you liked this, so why not try this approach’ but sometimes the use of AI in marketing can be misplaced.
A few weeks ago, a friend of mine had to put her cat to sleep. She’d spent a few months sharing pictures of him across social media and decided to announce his death on Instagram. Within seconds a comment popped up asking if she would like to “save the kitties”, her reply was surprisingly measured because she knew the comment was from a Bot, but it didn’t stop her from firing off an angry email to the company in question. The AI bot removed the human element of considering the context of the post and instead saw the tag #rescuecatsofinstagram and unwittingly pounced on her grief. The company never replied, by the way.
The use of social listening and sentiment analysis may have been a better choice for the situation listed above. Advances in AI mean that companies and brands can now analyse conversations on social media platforms, pick up issues and counteract them before they get out of hand. Two years ago, Samsung was able to pick up on a red tint problem on their latest S8 phone and fix it all through social media listening. They were also given the opportunity to strategically offer discounts to people who mentioned S8 problems. That intelligent solution was down to the brilliance of the human programming the AI bot.
AI in marketing does have its uses. Social media marketing has come a long way as a result of utilising technology to discover what users click on. The adverts that pop up on your Facebook wall don’t appear there by chance, but by careful targeting based on your preferences and things that you have clicked on. Where it doesn’t work, however, is when it is blithely used to target a broad audience with very few filters. I regularly have people asking me why, at the age of 44, they see adverts for incontinence pads and mobility aids. A closer inspection usually shows that the ad is either aimed at people over 35 or that they have an odd preference that has been logged as a result of clicking on something funny. On the one hand, then, yes, when combined with other brand interests, AI can be a fantastic social media marketing resource, but, yet again, there has to be a sensible human control.
As technology continues to reach heady new heights, Alibaba recently announced that they have developed a copywriting tool that can pass the Turing Test, we have to always keep in the back of our minds that it is still no replacement for employing a marketing professional. However advanced a chatbot may be, it still won’t be able to capture conversational nuances that the human eye and brain can. It won’t pick up on grief or upset. That’s why it’s still better to rely on the operator instead of the machine.
This blog was written by a breathing human being.
