How Supermarkets Use AI to Land More Products Into Your Basket
You know the story — you stop at a supermarket after work just to buy a carton of milk. Next thing you know, you’re leaving the store with a backpack full of groceries and $20 less to your name. And the weirdest thing is that you actually feel good about it. How do supermarkets always get you to buy more than you intended?
Grocers have many tricks up their sleeves; from sneaky shelf design that favors premium products to seductive snacks placed by the check-out queues to tempt customers while waiting in the line. But there is something that trumps them all — sales promotions.
There is something truly irresistible about promotions. The big percentage symbol, the “special offer” sign, the striking price tags in vivid colors — they all create a sense of urgency that makes you buy that box of cookies even though you haven’t planned on it.
But let’s be honest, we all love promotions. They help us to make faster decisions, discover new products, decrease the shopping time and find our favorite goods at a discount.
Finding a great deal feels like finding a gold nugget, but can be also equally tough. Our research reveals that on average, 97% of what is promoted is not relevant to any given consumer. So you need to go through hundreds of irrelevant offers to discover those few that are relevant to you.
And in today’s world where companies such as Amazon, Spotify and Netflix serve us highly personalised content with the tap of a finger, this can hardly be considered a great customer experience.
To make the bargain-hunting less demanding for the customer, the obvious solution is personalisation. But how to offer exclusive personalised offers to customers in a still predominantly offline world?
First, let’s look at an example — Amazon. The blockbuster deal to buy the U.S. chain Whole Foods for $13.4 billion in 2017 catapulted the ecommerce juggernaut near the top of the industry, sinking the stock value of many retailers basically overnight.
The reason Amazon is so dangerous to traditional retailers is because with its Prime membership, it has tons of data on hundreds of millions of consumers. This allows Amazon to target a huge chunk of the population with highly personalised promotional offers to drive them into Whole Foods stores on a daily basis.
The role of AI in the process
Amazon entering the market was a wake-up call for many retailers. Suddenly they started improving their loyalty programs and introducing new digital channels to improve customer experiences. Even discount chain Lidl has an app today.
Thanks to loyalty programs, brick-and-mortar retailers can already collect data and target customers on an individual level, but very little has been done in terms of personalisation. And this is where AI comes into the game.
The massive amount of transactional data that retailers collect is a true feast for machine learning algorithms. They can analyse hundreds of millions of shopping baskets and recognise complex market structures, consumers’ shopping behavior and common reactions to external stimuli such as discounts and promotions (including their form, time and placement). The algorithms can furthermore conduct thousands of micro-experiments and evaluate their effectiveness on the go.
How does it work?
AI is more than just a series of “if, then” rules, such as “If you buy pasta, you get a coupon for pasta sauce”. A true AI can recommend highly relevant products even though you have never bought them before (or at least not at this retailer). Basically, it might predict whether you will like a specific product before you even try it.
On top of that, AI can also predict exactly when you will need the product, and how much you are willing to spend for it. As a result, you’ll get the right offers at the right price and the right time. And you’ll get them via modern digital channels such as email, app, or a check-in coupon kiosk (a kiosk by the entrance to the store where you scan your loyalty card and get an instant print-out of exclusive offers).
These recommendations and discounts also save time. With around 40 to 50 thousand products available in an average supermarket, who wouldn’t struggle to make a decision sometimes? Getting personalised offers can help you decide where to go shopping and what to buy there, making your shopping trips quicker and more efficient.
So what’s there for retailers?
They get more sales because, thanks to a great customer experience, you chose them over the competition. On top of that, they can also earn additional revenue from CPG brands if they inspire you to try new products that you might like. You can learn more about how this process works in our article How Retail Media are Adapting to the Digital Age.
But it works. In a massive split test that we conducted with some of the leading German grocery retailers, our AI targeting system — the SO1 Engine — delivered nine times higher customer acceptance than general offers. That means nine times more people actually bought what was offered to them.
You might argue that the system could have cheated by offering bigger discounts to stimulate more sales, but in fact, the discounts were lower than those offered in control group of stores — by 36.4% . This shows two things — people value relevancy over the lowest price, and retailers can improve customer experience while also increase profits.
There’s only so much we can cover in one article about such a sophisticated technology. The way people are making everyday product choices is indeed a true science.