Retailers are deploying artificial intelligence today, and reaping the benefits. AI-driven demand forecasting can minimize inventory costs, while hyper-personalized digital marketing can improve return on digital ad spend.

Mylo Portas, head of retail, Peak

Mylo Portas, head of retail, Peak

As the dust settles on another peak season in retail, it’s a good time to take stock of where the industry is and where it’s headed. To make plans for the year ahead. To start—dare I say it—thinking about next Christmas…

In terms of where the industry is at, 2018 was a tumultuous year. Barely a week seemed to go by without another big name on the brink of collapse, and monthly statistics were used to paint a picture of the UK high street in perpetual decline.

But dig beneath the surface and a very different picture begins to emerge. Total retail growth for the year was steady and online spending excelled at an exceptional rate.

Simply being online, however, isn’t enough anymore. Consumers expect an ever more seamless, intuitive and personal omnichannel experience—and there’s one emerging technology that is making that possible from the smallest of retailers to the Amazons and Walmarts: artificial intelligence (AI).


A new approach to data

AI has been around since the 1950s, but has only began to gain real traction in business over recent years. As with anything that gains a bit of momentum, though, it’s become increasingly hyped, as people come to know the term without really understanding what it means or what the technology can do.

Far from tabloid sensationalism, AI is already having a positive impact in business. Our own retail industry report found that, on average,  e-commerce  companies using AI enjoyed 18  per cent  year-on-year  sales growth in 2017, 30  per cent  more than those that were not using AI.  The median gross profit margin for companies that use AI, meanwhile, was 50  per cent  higher than those who weren’t (48  per cent  vs 32  per cent.)

Becoming an AI-driven retailer has also led to a 75 per cent reduction in cost per social click during a recent trial.

So, while it’s clear that e-commerce businesses using AI have an edge over their competition, how are they actually implementing it?

On the front line

For retailers with in-store and ecommerce offerings, it’s important to create a highly personalised experience at every single touchpoint for customers.


Take leading fashion streetwear and sportswear brand Footasylum. It has seen a 28 per cent increase in email campaign revenue from hyper-personalised marketing communications since implementing AI into its ecommerce arm. Becoming an AI-driven retailer has also led to a 75 per cent reduction in cost per social click during a recent trial.

Regit—an online marketplace that allows consumers to buy, sell and maintain their cars—has also worked with Peak to become AI-driven. Using the technology to better predict when customers were going to change cars by creating machine learning models meant the company was able to create a lead score that was then assigned to customers in its CRM. This helped its call centre agents to better personalise the service they gave to customers, based on how likely to model predicted they were to change their car.

The result has been a 27 per cent increase in sales, while also reducing operational costs by 35 per cent. This is due to the fact that Regit was able to better plan the number of staff it would need in the call centre at certain times of the day, maximising productivity without having to increase the number of employees overall.

That’s where AI has helped on the front line. However, it can also be applied to the bane of any retailer’s life: stock levels.


Stocking what you know you need, not what you think you need

Keeping stock levels at the right levels can be a challenge. Too much ties up cash and space (which requires more cash), while too little can lead to shortages and a poor customer experience. Some retailers take a ‘just-in-time’ approach, a bit like manufacturers, to limit the amount of storage space they require. This is fine, yet does expose them to supply chain failure risk, particularly at busy times of the year.

AI can help retailers manage their inventory more effectively by improving their demand forecasting. By giving them insights into what customers want, and when they want it, ecommerce businesses are better able to anticipate demand and optimise their inventory accordingly. This goes beyond having Christmas decorations in stock in the final months of the year—it’s about knowing how many Christmas decorations are going to be bought each week in the lead up to 25th of December and bringing in stock accordingly.

Our report found that, when companies use AI, the median number of stock days was 43—a 9  per cent  reduction on the 47 median stock days of companies not using AI. That’s less time in storage and quicker realisation of revenue. While it may seem like a small difference, add it all up and it starts to have a significant impact. At a sector level, our research found that food and beverages retailers can reduce their inventories by 1.4 billion pounds by implementing AI, while the automotive sector could free up 4.7 billion pounds.


The future is here

AI has the power to help ecommerce businesses perform better. It’s no longer the stuff of science fiction; it’s a real technology that forward-thinking retailers are already using to great effect.

They are harnessing their data to develop stronger relationships with their customers, based on personalised experiences that generate repeat business. For many, AI in retail is still in its early stages, and it offers businesses a way to secure a competitive advantage. However, it will soon be a part of everyday business, and only those that integrate it into their operations now will accelerate away from the pack.

Peak is a UK firm that specializes in artificial intelligence.