Artificial Intelligence is making marketing smarter. Here are 5 AI capabilities you need to know about.

Last month, an Elon Musk-backed company revealed it had created an AI text generator so revolutionary that they are refusing to make their research public for fear the technology could be misused. Or perhaps lead to the rise of the machines and we all know how that ends up...

Marketing AI isn’t up there to the Elon Musk and Skynet levels of sophistication, but it’s amazingly impressive nonetheless. And as the tech improves, smarter digital marketing opportunities become available. Let’s take a look at five that are already making waves.

 

1. Search engines

We take search engines for granted these days. Not only can we search in a myriad of ways (voice, image, good old fashioned typed words), but also the quality of the search experience has improved dramatically.

This is in no small part due to Google’s AI, RankBrain. Introduced in 2015, it’s a machine learning-based algorithm that allows Google to retrieve smarter results ­– like auto-correcting mistakes, auto-filling, using natural language and semantic searching.

Machine learning in search AI has been so successful that now most major retailer sites have incorporated it into their search engines, allowing retailers to suggest relevant items to purchase rather than the shopper having to know precisely what they want.

2. Dynamic pricing

So your search engine has found what you’re looking for, but is it at the right price? Machine learning can now retrieve and analyse a shopper’s buying habits and predict how much they would be willing to pay for an item and their propensity to be swayed by a special offer or discount.

What this means in practice is that businesses can use dynamic pricing to not only precisely target customers, but also know just how much of a discount they’d need to place on a product to get the sale over the line.

It’s basically an AI version of The Price is Right.

3. Smart recommendations

We can all relate to this: endlessly browsing through Netflix, trying to decide on what to watch from the seemingly infinite choice of shows. But according to Netflix’s own research, actually that engagement time is significantly shorter. According to their studies, we give up between 60-90 seconds of browsing and then get distracted by our phones, life or that penny on the floor.

So with such a small window to keep the viewer watching, how does Netflix retain our attention?

You guessed it – more AI algorithms to make highly personalised recommendations. This isn’t just based on what you last watched, but a whole host of additional data – including personal profile and audience demographic. By getting to know you using your data, brands can recommend more relevant products or content.

It’s all been quite passive up to now, hasn’t it? These algorithms silently running in the background without you knowing. But Sky is inviting viewers to the party, by actively asking them what mood they’re in and then suggesting content. By asking specific questions like these, the recommendations can be honed to an even more accurate degree to keep the viewer engaged.

4. Visual search

The most recent advancement in search engine capability, and perhaps the most exciting for retail marketing, is image recognition and recognition.

For retail, it opens up huge possibilities to personalise the shopping experience. Fashion is an obvious example. Businesses can recommend items of clothing not just based on previous shopping behaviour, but how the garments look ­­– matching a shopper’s unique individual style with similar products available.

Fashion retailer Asos has developed a platform to do just this. Using its Style Match tool, shoppers can upload an image and its AI visual search algorithm then finds similar looking items in the Asos catalogue.

London startup Thread uses a similar mechanic. Upon visiting the online shop, customers sign up, upload a few photographs of themselves, what's currently in their wardrobe and their measurements, and then they are paired with a virtual personal stylist - a human faced AI that suggests looks and products based on your profile data. 

‘Shop the look’ is going to be one retail phrase you’re going to hear much more of.

5. Sentiment analysis

One of the most targeted – if utterly creepy – use of an AI algorithm is how brands use it to “listen” to conversations on social media and then analyse the sentiment. If negative attitudes are being conveyed about the brand, the AI can quickly identify them and counteract them with a response.

It may also be used to look out for buyer’s intent. If someone on social was asking for opinions on that brand’s product, or even a similar competitor’s one, then that interaction would tell the AI to target that person with a relevant ad or dynamic price.

 

This “social listening” begs the question of how far an AI can go without customers feeling their privacy is being invaded. Yes, the use of AI is undoubtedly making our brand experiences better, more streamlined and more fulfilling, but is there a point where automated marketing and learning algorithms can become victim to their own success? Does familiarity breed contempt?

 


Picture of Alex Allston, Senior Copywriter at The Sharp Agency

Alex Allston

Senior Copywriter