The Role of AI Technology in Solving the Problem of Food and Beverage Wastage

Image Description
TrueStock Team October 5, 2020

The Problem of Food and Beverage wastage

Food and beverage businesses can handle many hundreds of products and must determine the stock for each product daily. Many of these products may be perishable or have limited shelf lives due to expiration dates.

Food and beverage wastage and price reductions due to imminent ‘use by’ or ‘best before’ dates can seriously impact the amount of product that must be disposed of.

In the fast-moving sectors such as hospitality and food service, which combines both production and sales, food and beverage wastage can be high due to inaccuracies in demand forecasting and subsequent procurement of food and beverage materials.

Over ordering stock creates waste and running out of menu items can lead to unhappy customers.

Seasonal changes in consumer demand is a key variable that food and beverage sectors must consider when planning food items to be produced or procured e.g. ice cream in the summer and food or drinks specific to the Christmas season.

The weather can have a major effect with dramatic weather changes e.g. flash snowstorms or hot weather impacting the demand for hot or cold foods or beverages.

Nearby events can have an impact on retail food outlets and restaurants. If there is a concert, festival or a sporting event food and drink is very often top of mind for large groups of people attending events.

The cost of food waste in the retail, hospitality & food service and manufacturing sectors are significant within the UK.

The value of food waste in the UK
Source: IDG

WRAP (Waste and Resources Action Programme) was created in 2000. One of their aims is to help organisations to unlock the economic benefits of waste reduction and resource efficiency in the food and drink supply chain.

Some of the key issues that WRAP is trying to address is the costs of wasted food and the impact on the environment.

In 2018 the IGD and WRAP launched and a new industry Food Waste Reduction Roadmap and have commitment from some of the UK’s largest retailers, food producers and manufacturers, and hospitality and food services companies.

Research conducted by WRAP and the World Resources Institute found that there was a strong business case for reducing food loss and waste with the average business achieving a 14:1 positive return on investment in waste prevention measures.

Food waste producing sectors in Ireland
Source: has identified some of the top food waste producing sectors in Ireland. Stop Food Waste is a programme funded under the EPA National Waste Prevention Programme.

Their emphasis is on waste prevention as food waste in the different sectors can be avoidable. They claim that poor demand forecasting is a significant reason for wastage.

Example case of inaccurate demand forecast

As a convenience store manager and you need to determine how much strawberries your store needs for the next week to order in. You look at your historical sales and calculate an average of how much you think you need. The order gets created and you wait.

Your order gets delivered but the weather gets worse for your store location, sales of strawberries and cream start to reduce. Your store now has an excess of strawberries that are now going to waste.

Strawberries are just one of the thousands of food products that are wasted each year. The weather is not the only factor but one of the biggest impactors of retail sales demand.

Example case of costs to a restaurant

If you have a restaurant with 9 full time equivalent staff, then you could be generating:

9 x 0.8 = 7.2 tonnes of food waste annually

Taking the cost of food waste at €3,000 per tonne, the cost to this business can be estimated at €3,000 x 7.2 = €21,000 per year

Tackling the Problem with AI Machine Learning Technology

The challenge to most companies is dealing with the large amount of food and beverage product data and demand driving internal and external factors. Predicting changes in demand to enable a business to make informed decisions about future sales and purchasing.

Our machine learning solution provides those businesses who work in the food and drink sector to accurately produce demand forecasts. Data to include sales history along with other factors such as the weather, events, promotional periods, seasonality and many more.

Typically, other methods of sales forecasting for example moving averages do not take into consideration factors such as the weather, events seasonality etc. and consequently inaccuracies in the sales forecast can result in product wastage.

"With machine-learning technology, retailers can address the common and costly problem of having too much or too little fresh food in stock."

Utilizing learning algorithms, large amounts of data and integrated external data sources that impact sales demand can be analyzed with speed.

Constantly and automatically improving with time, machine learning enables companies to respond quickly to fast changing business situations.

Machine learning algorithms help retailers determine demand
Source: McKinsey & Company

As Truestock’s sales forecasts for products are tailored specifically to each store location, this increases the degree of accuracy for the forecast.

Being able to predict sales and stock replenishment requirements more accurately food wastage can be prevented, and significant cost savings can be made.

Find out how quickly you can generate a forecast

When TrueStock generates a forecast it provides a daily breakdown of how many sales are expected. This is done in the form of percentiles, P10 is the lower estimate, P50 is the middle estimate and P90 is the upper estimate.

This forecast range helps make more informed decisions for your business.


Get started with up to 10 variant forecasts for free forever. Paid plans start at $25/month.

Free To Install
Image Description