OPINION | Could AI-driven forecasting be the surprise winner this summer?

Traditional demand forecasting in retail was built for a stable world. However, major sporting events are anything but stable. Jaime Silvester, EVP of Retail, Circana, argues that this makes a bumper summer of football the ultimate stress test for AI‑driven forecasting and that lessons learned from AI‑driven forecasting could shape how retailers operate in the future.

The games are in full swing as consumers experience a mix of national pride, excitement and drama over the next few weeks. Later kick-off times for fans in England, Scotland – and across Europe – means more socialising and shared experiences as friends and family enjoy the action together. 

Naturally, this fuels last‑minute and late-night shopping and increased demand for food, drink and party essentials, particularly at weekends, when households gather to recreate the match‑day atmosphere at home.

All of this is set to deliver a boost to the retail sector. This is despite the latest Circana data indicating that households may be trading down to cheaper products or buying on promotion due to the cost-of-living crisis. 

If sales from the Euro 2024 are anything to go, we can safely assume that with both England and Scotland set to go through to the knockout stages, fans will be celebrating with beer and snacks.

Predicting the unpredictable

Unfortunately, like the scores at many major sporting events, sales forecasting can be just as unpredictable. The lessons learned from four years ago – or even two years ago – may not necessarily apply today.

In fact, retailers face one of the most unpredictable periods of the decade. From upsets and disappointments to peaks and troughs in national pride, this summer’s football fever is on a different scale; creating demand patterns that historical data alone cannot predict.

With billions of viewers worldwide, the tournament consistently remodels how, when and what consumers buy – and not just die‑hard football fans, but those who may not be fans, but live within households shaped by football routines and purchasing behaviour. 

Powering demand

Retailers using AI in demand forecasting can anticipate and respond to rapidly shifting consumer demand as the tournament unfolds. 

By linking forecasting models to real‑time signals, such as match times, public screenings, transport flows (or traffic reports) and even weather, retailers can align their inventory with where fans are, not where last year’s data says they should be.

For example, with fan zones and screening events in many towns and cities throughout the country set to screen the major matches, AI models can help retailers understand how demand moves along commuter routes and local areas before, during, and after kick‑off. 

This ensures availability of high‑demand products at the right time, in the right locations.

Similarly, late night kick-off times will provide smaller stores and those in major transport hubs like railway stations with a much-needed spending boost particularly during the working week when commuters are travelling home to watch the game.

Using AI forecasting, retailers can adjust ranges – from food and drink pairings to promotional offers – dynamically throughout the tournament. Traditional match‑night staples like snacks, beer and pizza can be placed next to adjacent categories such as BBQ favourites and party staples. 

Just as success for a team drives spending, when teams exit earlier than expected, demand will drop suddenly, leaving retailers with the headache of dealing with unsold stock. AI allows retailers to remain agile, which should limit overexposing themselves.

The ability to connect inventory, location, and shopper behaviour is critical in redistributing stock, as well as creating opportunities to target promotions or deals locally – and, most importantly, reducing waste.

It is important to remember, however, that AI is not a replacement for retail judgement – and should never be treated as such.

The most successful AI strategies combine machine intelligence with human oversight, using AI to surface patterns and risks, while people make crucial judgment calls. The human element ensures accountability and sensitivity to the context within which the AI data is utilised.

A summer of football mania presents a rare and exciting opportunity for retailers. Emotions run high, behaviours change and fans respond in very different ways. It may only last a few weeks, but AI‑driven forecasting could be the real winner of the tournament – and the lessons learned could shape how retailers operate long after the tournament ends.