Artificial intelligence in retail

Graham Lack

18th Aug 2017

Are big data, cognitive learning and neural networks the buzz words of our time or is there something actually in AI that can help retailers in the twenty first century?

Clearly we are in the information age. We are surrounded by it and interacting it with it every day. As you start your car on a Monday morning, your phone will tell you that you’ve got 35 minutes travel time to work!

Moore’s law refers to an observation made by Intel co-founder Gordon Moore in 1965. He noticed that the number of transistors per square inch on integrated circuits had doubled every year since their invention. Moore’s law predicts that this trend will continue into the foreseeable future. Note the date. I wonder if he thought 2017 was the “foreseeable future”, but irrespective of that, the law has stood the test of time and if researchers solve the quantum computing problems will continue to do so (albeit nothing to do with transistors on circuit boards).

As computer power increases, so too do the critical algorithms needed to run them. In the past, algorithms told the computer “how to think”, but now it is more about teaching the computer “how to learn”. AI is real.

As much as I love technology, I’m struggling with AI in retail. I read an article recently that left me a little underwhelmed. Here are some examples:

Case 1. Pepper Robot is a humanoid robot that can interact with customers and “perceive human emotions.” The article gives an example where, on Pepper’s introduction, “the retailer experienced a 98 per cent increase in customer interactions, a 20 per cent increase in foot traffic and a 300 per cent increase in revenue”.

Really? Even if the numbers were true, over what period of time were these values sustained? Are we talking a year or the day Pepper arrived in-store, no doubt among much fanfare. Be wary of quoted statistics. Furthermore, was it the AI or the novelty factor that drove any alleged increase? Life size Ronald McDonald statues probably had similar impacts. Not much sign of AI there.

Case 2. Outdoor retailer “North Face adopted IBM Watson’s cognitive computing technology to help consumers determine what jacket is best for them, based on variables like location and gender. For example, hiking in Iceland in October and commuting in Toronto in January will yield different results”.

Once again, you don’t need much AI (certainly not the power of AI that IBM’s Watson is capable of) to work out that different climates means different jackets. Isn’t that why the buyers at North Face put multiple jackets into the range in the first place?

Case 3. Amazon Go trialled the employment “of check-out-free technology that allow customers to shop and leave. Customers use the Amazon Go app to check in, but thereafter the entire shopping experience is designed to be automated. Sensors track which objects customers pick up and put in their basket, and customers’Amazon accounts are automatically charged after exiting the store”.

As clever as the concept is, this is barely an example of AI. Very techie, but really just data capture, presumably with RFID tags.

To be fair to the article, the author also questioned the validity and relevance of the case studies and like me, concluded, that if AI is going to make headway into retail, it is mostly likely going to be back-office functions and logistics where the benefits can be rigorously tested.

The so called soft-benefits, like improved customer experience, should always be taken with a grain of salt.

This story first appeared on our sister site, Inside Retail Australia.