Coles to use AI and machine learning platform to predict liquor demand

Coles will use AI and machine learning to make better predictions of liquor consumption including weather forecasts to try and make sure bottle shops are stocked for key events.

Coles will harness AI and machine learning to anticipate liquor sales. Picture: David Crosling

Coles will harness the dazzling possibilities of AI, machine learning and algorithms to predict consumer demand for beer, wine and spirits sold by its liquor chains, helping it to alert suppliers in advance and boost the efficiency of its supply chain running from the vineyard, brewery and distillery to the store.

Fed with data around key annual events as well as weather patterns, it is a brave step into the new world of AI to anticipate consumer tastes and preferences around liquor which can be a highly fashionable or trendy category that can quickly change — with a bout of hot weather pushing drinkers to prosecco or a winter chill at Mother’s Day pushing sales of Irish cream liqueur.

The supermarket chain, which owns liquor retail banners Liquorland, First Choice and Vintage Cellars, has begun introducing an advanced forecasting and automated ordering system for its liquor suppliers built to fit the unique needs of its business and has begun shifting liquor suppliers onto the system.

This AI and machine learning platform has been in a trial phase for six months which the retailer has viewed as successful, and now Coles Liquor is commencing its phased rollout of the forecasting system for suppliers which will continue into 2025.

It is hopped internally at Coles the state-of-the-art AI and technology kit will help better forecast the needs of drinkers, especially around key events such as New Year’s Eve or Christmas, and then enable the supermarket’s liquor stores to order in advance from suppliers to meet that expected demand.

The algorithm being used by Coles Liquor takes into account a variety of factors including events, campaigns, promotions, past sales, and weather to generate a demand forecast. The demand forecast is then fed into Coles Liquor’s ordering system, to create an order plan for the business and a purchase order for the supplier.

It is hoped this will give suppliers access to predicted orders which will help reduce uncertainty for them in the ordering process. This will help them plan manufacturing and procurement in advance.

“Our forecasting and automated ordering system will help us predict customer demand and required stock levels in advance,” a spokesman for Coles Liquor said.

“This allows us to effectively tailor our range to local communities and better replenish stock, meaning our customers can always rely on us to have their favourite drinks on our shelves, where and when they want it.”

For example, a supplier who has a line of sparkling wine would have a more accurate forward indication of which stores are likely to need certain quantities, at what times of year, like Mother Day, New Year’s Eve, and it would also take into account where the line has been on promotion in the past, or any impact to sales where there have been adverse weather conditions in the past.

“By having access to better data and insights we can effectively collaborate with our more than 2000 suppliers and give them greater visibility of stock requirements in advance to support their production planning. This is particularly important for our small and medium suppliers so they can ensure availability of their stock in our stores at the right place at the right time.”

Three years ago Coles took on a similar AI platform to replace manual ordering for the replenishment of its fruit and vegetables for its supermarkets, and companies are increasingly turning to AI and machine learning to streamline their operations.

Originally published as Coles to use AI and machine learning platform to predict liquor demand