May 6, 2024

Markdown analytics to reduce food waste at grocery stores

A case study on improving pricing strategies

Markdown analytics to reduce food waste at grocery stores

At a glance

Challenge

A large Portuguese food retailer was facing a common problem in the industry: food waste due to expired products. With the goal of reducing food loss and waste the retailer turn to analytics and partnered with LTP.

To mitigate the food waste, the retailer places a new tag with a reduced price for each item approaching its expiration date. The challenge was to dynamically set this markdown price in order to promote consumption and avoid food waste.

Solution

We developed a data-driven approach using sales history, markdown decisions, stock, and product/store attributes to recommend optimal discounts. This method balances competitive pricing for consumers with maintaining key retail indicators. It evaluates sales volume and margin impact, demand cannibalization, and potential losses from expired products. A core element is a model that predicts sales probabilities for nearly expired items, factoring in markdowns, timing, promotions, and expiration proximity.

Results

The new markdown approach improved process performance by 20 percentage points, reducing food waste and increasing sales. It also eased store clerks' workload and enhanced process oversight. Additionally, the data-driven system provided valuable insights into pricing strategies and product sales performance, paving the way for advanced research into price elasticity and improved consumer service.

Challenge

A large Portuguese food retailer was facing a common problem in the industry: food waste due to expired products. With the goal of reducing food loss and waste the retailer turn to analytics and partnered with LTP.

To mitigate the food waste, the retailer places a new tag with a reduced price for each item approaching its expiration date. The challenge was to dynamically set this markdown price in order to promote consumption and avoid food waste.

Approach

Solution

We developed a data-driven approach using sales history, markdown decisions, stock, and product/store attributes to recommend optimal discounts. This method balances competitive pricing for consumers with maintaining key retail indicators. It evaluates sales volume and margin impact, demand cannibalization, and potential losses from expired products. A core element is a model that predicts sales probabilities for nearly expired items, factoring in markdowns, timing, promotions, and expiration proximity.

Results

The new markdown approach improved process performance by 20 percentage points, reducing food waste and increasing sales. It also eased store clerks' workload and enhanced process oversight. Additionally, the data-driven system provided valuable insights into pricing strategies and product sales performance, paving the way for advanced research into price elasticity and improved consumer service.

Our
AI-generated
summary

Our AI-generated summary

Our AI-generated summary

A large Portuguese food retailer was facing a common problem in the industry: food waste due to expired products. With the goal of reducing food loss and waste the retailer turn to analytics and partnered with LTP.

To mitigate the food waste, the retailer places a new tag with a reduced price for each item approaching its expiration date. The challenge was to dynamically set this markdown price in order to promote consumption and avoid food waste.

We developed an approach that leverages existing sales history, past markdown decisions, stock positions and product and store attributes, allowing managers to establish different strategies on a product level. The approach considers the product’s financial and performance data and stores’ operational features, to determine the recommended discount.

Our AI-generated summary

Our AI-generated summary

Three relevant indicators were considered to determine the best markdown price: the impact on sales volume and margin; the effect on demand for other items; and the expected losses if the product reaches the expiration date.

One of the key ingredients of the approach is a model that estimates the sales probability of close to expired items of each product, in each store, considering numerous features such as markdown price, weekday, promotional intensity, number of days to expiration, among others.

The implementation of the new approach consistently increased the performance of the markdown process by 20 percentage points.

It promotes products consumption, by offering a more competitive price to the consumer while keeping track of retailer’s main indicators.

This resulted in a direct reduction of food waste and sales increase.

Furthermore, the new data-driven process also substantially reduced the effort of store clerks when setting the new price and boosted the overall control of the process performance.

An important byproduct of the implemented novel analytical process is the valuable insights into how the markdown process and other variables impact the products’ sales performance, broadening the scope of understanding in how pricing strategies outside the scope of near-expiration products can be improved: the data collected from these new data points supports research and methodologies regarding price elasticity, which can be further developed to better understand and serve the consumers.

Read more

Read more