A sneak peek into AI’s weird, wacky and wonderful consumer insights


December 7, 2018 11:46 am

Did you know that viewers of the American fantasy drama television series, Game of Thrones, are highly likely to be the same people who purchase sex toys online, particularly on a Monday? Would it ever occur to you that online shoppers who buy used car parts are nearly always also interested in used bottles of perfume? What about that, between 8pm and 11pm on Thursdays, Texas residents who are searching for robot pool cleaners will undoubtedly end up purchasing an item of that description, sometimes even more than one?

These are some of the most out-of-the-box, but incredibly accurate (nearing 100%), insights obtained in the application of an artificial intelligence (AI) consumer behaviour prediction algorithm.

While these insights are interesting, embarrassingly relatable or just plain hilarious at first, they are extremely valuable to businesses operating in the relevant industries. These are insights bringing completely unknown demographic connections to light, connections that no human would ever think to look for.

Applying this knowledge to existing strategies

How do these more peculiar connections help businesses? This consumer prediction algorithm matches people with products, inventory with opportunity, price with spending propensity and people with usage patterns. These insights shed light on how specific demographics shop, when they shop, what they buy, how they engage and what they will most likely want to buy in future. By applying this knowledge to existing strategies, businesses can set themselves up to out-predict their competition, maximise their efficiency and enhance their customer satisfaction.

So, for the over $15bn sex toy industry, the Game of Thrones insight would be really powerful in planning moving forward, from how they engage with customers and the kinds of products they look to stock, to where and how they market these – all to ultimately boost their bottom line.

The same can be said for those dealing in robot pool cleaners in the Texas region. Increasing marketing efforts during a particular time slot with such a high sales conversion rate would be massively beneficial for business.

In the local context, the algorithm has identified a significant, and fast-growing, market segment which has been largely untapped by businesses across industries – African expats living in South Africa.

Again, this is a demographic that the algorithm found without anyone thinking to look for it. What’s more, the algorithm picked up a common consumption pattern for people within this demographic – items that are Islam-centric in nature. Since people following Islam are such a small minority in South Africa, businesses do not necessarily have a customised offering for them.

Using existing data to make these connections, and have real insights on existing consumers, businesses in entertainment, hospitality and retail – to name just a few – could adapt their offering to better cater to this market which follows a different weekday setup with weekends falling on Friday and Saturday.

AI has the power to find connections

Media entertainment providers might also want to re-evaluate their programming to offer suitable weekend content on a Friday and restaurants might look to optimise their share of this market by introducing family specials for a Thursday evening. But they could only take these strategic steps if they know that the market exists, which they wouldn’t if it wasn’t for AI.

While conducting focus groups is a traditional method for understanding the market of a particular business, this is a time-consuming exercise with a number of limitations from sample size, to an emphasis naturally being placed on known, existing demographics. The speed at which AI can pick this up far exceeds any other approach, providing accurate predictions within as little as just two weeks of receiving the data.

I often share these insights with top business executives looking to embrace the power of AI in consumer prediction, they are evidence that there really is no such thing as useless data.

AI has the power to find connections that may not naturally exist but which hold unlimited potential for profit, business growth and customer satisfaction.

Categorised in: