Thought for Food: How Can AI Help Supermarkets Innovate Their Fresh Food Replenishment Model?

IN — 25 September, 2017

As a consumer there’s nothing more frustrating than dashing to the supermarket on your way home from work, only to find that the specific item you want is out of stock. Whether it’s fresh herbs for a recipe you’re cooking that night, or burgers for the impromptu ‘sunny day’ barbecue you are hosting, being greeted by an empty shelf can ruin your plans.


On the flip side, it’s an exciting bonus when you head into the supermarket to pick up a few groceries and swing by the ‘reduced to clear’ section – only to find it stacked with enough almost-out-of-date goods to fill up your freezer. Great for me, not so good for the retailer’s profits.

These contrasting scenarios might evoke very different reactions among consumers, but for retailers they have two things in common:

1 – They happen every day at supermarkets everywhere around the world

2 – They are both bad news for the bottom line

Turning complexity into opportunity

The complexities of fresh food replenishment are well known by the grocery industry, but for most supermarkets, a solution hasn’t yet been found – resulting in the UK’s biggest supermarkets throwing away or donating over 200,000 tons of surplus food each year.

The answer to enhancing fresh food replenishment lies in finding an intuitive technology to manage the fresh category, and to optimize replenishment decisions long term. And the most effective solution is Artificial Intelligence (AI).

From fixed rules to fluid agility

One of the reasons that AI is being put forward as a solution for fresh food replenishment is its ability to go beyond data analytics and offer actionable change. Sophisticated algorithms will enable interpretation of multiple data sets surrounding the consumption of fresh produce, which can then be intelligently converted into decisions about how much stock to buy, in what location, and how often.

Up until now, most replenishment and forecasting models have been based on fixed rules, which cannot necessarily accommodate unknown influences on consumer behavior. From a sudden heatwave to a major television cookery show featuring a certain recipe, there are thousands of factors that can shape when and why people buy certain products. And with its limited shelf-life in fresh food, this sector can present the biggest opportunity for winning or losing the margin contest. Too much and the reduced to clear section is stacked high once more; too little and potential customers go home empty handed.

Automatic for the people

Given these complexities, AI can provide grocery retailers with a more flexible way to manage fresh food replenishment. These sophisticated algorithms we’ve already mentioned are capable of not only processing a much greater volume of data than even the most talented human being, and at much greater speed; they can model many more possible scenarios, to provide retailers with an accurate forecast based on what is actually happening at that point in time.

Moreover, the right AI solution will be able to automate decisions as a result of these scenarios, to set off a chain reaction that ensures the right number of products are dispatched to each and every store, with minimal risk of overstocking.

Not only is this a more efficient and effective way to manage fresh food replenishment, but AI solutions are also self-learning, which means they grow stronger over time. The intelligence and intuitiveness of the technology develops continually, to keep refining the fresh food experience around customers’ ever-evolving needs.

For a round-up of retail AI innovation visit our AI in Leadership page – where you can read the latest articles from The Daily AI.

Blue Yonder Blue Yonder

We enable retailers, consumer products and other companies to take a transformative approach to their core processes, automating complex decisions that deliver higher profits and customer value using artificial intelligence (AI).