Close this search box.

How to Reduce Waste on Grocery Shelves

Retailers have been struggling to keep up the pace with shoppers’ changing needs in times of the COVID-19 pandemic. Fresh produce waste has increased significantly from 2-3% to as much as 10-15%.

Learn how to save your razor-thin margins by optimising product allocation and adjusting prices as needed. This two-fold approach will maximise your profits, reduce costs associated with spoilt goods, and allow you to run your business sustainably.


Product Allocation Under Control

The frequency of customers’ shopping trips has decreased in favour of buying more items in bulk — wreaking havoc on consumption patterns and trends. Reduced sales of fresh, ready-to-eat and ready-to-heat products are a real challenge for grocery retailers, as they must clear the shelves of such products before they expire.

The extraordinary fluctuation in demand was a significant issue for those in charge of supply chain management in the early days of the pandemic. Entire boxes of shelf-stable products were bought out while the sale of fresh items decreased. Combined with the significantly higher number of eCommerce orders, the in-store stock replenishment team is at risk of running out of capacity to fulfil their daily duties. The allocation and replenishment systems during this period don’t work in the most optimal manner. They are based on historical data, which needs time to accommodate to the rapidly changing circumstances resulting from the pandemic. Commercial-off-the-shelf (COTS) solutions tend to have limited capabilities and offer generic user experience with little or no possibility for customisation.

At Objectivity, we helped one of our clients by providing bespoke software as an intermediary component between the demand forecasting COTS system and a legacy tool for stock and ordering. The store manager can manually override volumes or the frequency of automatic allocation results if they do not match the current demand, workload, or stock capacity.

For example, warehouse workers were not able to handle the increased volume of inbound deliveries, even after hiring new employees. Our application allowed management to change the set of allocation rules for non-essential products from daily to weekly to reduce the workload in the stockroom. Shelf-stable food was restocked less frequently, but in larger quantities, which reduced the effort of shop floor employees. These changes permitted the staff to focus on handling fresh products, and to increase the amount of time they are displayed on store shelves, thereby maximising the sales window and reducing waste.


Data Science Comes to the Rescue

COVID-19 exacerbated waste issues, however this matter has been posing challenges for retailers since before the pandemic. This is likely because COTS products are, by default, designed to handle generic processes.

Our Data Science team analysed one of our client’s retail operations by auditing product allocation results to identify inaccurate sales forecasts. As a result, the supermarket chain discovered significant inconsistencies between forecasts and actual sales data.

Over 25% of goods had a difference of over 8% between actual and forecasted sales results. Most of the products with the highest discrepancy were in the fresh produce category. Having access to this type of data can enable retailers to optimise costs. The mean absolute percentage error measure (MAPE) was off the charts for some fruit — especially given that the expected result was to stay below 25%. As such, it turned out that much less stock was sold than expected.

A detailed audit of the retailer’s operations showed that forecasts were accurate in the past and started to deteriorate over time. The reason for the issues was quite surprising. It turned out that the forecast was stable, but there were discrepancies between the stock balance in the system and actual stock on the shelf. Products were missing, and they had to be replenished, even if the inventory was accurate. The system incorrectly assumed that there was plenty of stock, leading to lost sales and a disappointing user experience caused by the unavailability of stock or the inability to find the desired product.

Our Data Science team can challenge the COTS package to identify and address gaps or exceptions in forecasting. Either by analysing the combined sales, waste, and inventory allocation data or by investigating the logged occurrences of the manual amendments by store managers.

However, even the best product allocation tool will not provide a flawless forecast. There will always be certain discrepancies. Improvements in this part of the supply chain can help reduce the volume of issues — nevertheless, they must be paired with an in-store solution.

Effortless Signage

In today’s fast-paced world, consumers don’t have time to look for a carefully hidden expiration date printed in the smallest possible font size. They storm through the grocery store, grabbing the products they need and expecting them to be fresh.

A few days ago, I bought a salad and only noticed it was expired after I got home. I was disappointed and felt a little cheated. I wasted money on a useless product. I couldn’t eat what I planned. I needed to throw away the food — which is usually a last resort, as I’m conscious about the environmental footprint of the manufacturing of goods. So, I decided that the next time I need to go shopping, I’ll go to another retailer, despite the fact that the other store is located further from my apartment. I ended up following through with my plan.

Retailers need to meet consumers’ growing quality demands (especially when it comes to fresh goods) and be consistent with legal requirements. Fresh produce maintenance is a real challenge because of a lack of widely used industry standards for ensuring traceability and recall. The GS1 DataBars, which allow for the automatic management of expiration dates, aren’t implemented by all suppliers. Most of them still rely on EAN13 GTIN indicators, which only provide generic information. In an attempt to get ahead of the competition, retailers often choose to develop custom solutions aimed at helping them to quickly sell products at the cusp of expiration.

The lack of an efficient way to automatically detect the best by date shifts the focus onto improving store rotation processes in a way that will enable perishable items’ prices to be reduced before they expire. Time is a crucial factor, as with every hour the value of fresh goods drops significantly. They can become spoilt even before the markdown tags are applied. If you can’t fully automate the process, make it the easiest it can possibly be by utilising industrial barcode scanners integrated with a portable label printer. This duo of handheld devices will enable you to generate shelf edge labels ad hoc, without having to take a time-consuming stroll to the back office in search of supplies. Moreover, the reduced price needs to be synchronised with and recognised by the checkout system to avoid discrepancies, which can lead to an increased number of complaints. Fortunately, the industrial barcode scanner and portable label printer are network-connected devices, so the price is updated on the system’s backend in real-time.

Applying a clearance tag to an item increases the product’s chances of being purchased. However, this isn’t as easy as it sounds. The retailer may be in breach of the Consumer Protection Law if the updated label does not meet a set of tangled legal regulations, which vary for ambient and fresh products. When we worked with one of our clients on the development of a set of business rules meant to determine whether price reductions could be visualised in the form of a cross-out price, we encountered many challenges. Having worked on such a project has taught us how important it is to have an in-depth understanding of regulatory compliance signage.

Once you have your network-compatible devices ready-to-go and you’re all caught up with legal regulations, the last step is to ensure that the shop floor employees place the clearance tags on the products. Workforce organisation is challenging especially for small or discount retailers, where the store manager is overwhelmed with the number of tasks. One of our clients approached us with this very problem. In an answer to the challenges they were facing, we developed a system capable of automatically creating tasks for store staff. The application enables central or store managers to monitor the progress on current data, as well as browse historical data to assess employees’ performance. This improvement led to a significant reduction in overdue or missed assignments, as the visibility of in-store operations can now be audited from the client’s headquarters. This supports remote work and is especially important in today’s uncertain times when the staff located in the headquarters can’t visit stores as frequently as before.

Some retailers are taking advantage of dynamic product labelling via the internet which seems to be the next big thing when it comes signage processes. At first glance, this type of solution seems amazing — a remedy for all the problems mentioned above. Unfortunately, upon closer examination, a series of economic and technical blockers start to appear. Battery life and reliability issues can be a nightmare. Electronic label vendors are making the point that standard processes are labour-intensive and error-prone. However, they tend to forget that human interaction can solve multiple issues not addressed by their solution. Employees can review if the price and product description is accurate, check the expiration date, validate if the right product is on the right shelf, report and address gaps, and update stock count. All these factors combined with the high cost of the introduction and maintenance of Electronic Shelf Labels render them a technology innovation that is likely to become popular in the future. However, at the time of the COVID-19 outbreak, your business might have to meet certain urgent needs with the use of more agile and tactical solutions.


Accurate and flexible product allocation supported by data science and lean in-store operations for markdowns can help to significantly reduce waste. By optimising business processes around fresh produce, retailers can sell more products and help to position themselves as an environmentally conscious company. Sustainability is something that many of today’s consumers truly care about. Customers are ready to pay extra for goods if they’re sold by a company that cares about the future of our planet. Fortunately, Artificial Intelligence techniques and portable handheld devices (connected to your core systems) can be effective at solving one of the retail industry’s key challenges — waste production.

About the Author

Paweł Krawczyk

Chapter Leader & Senior Software Developer at Objectivity. He is passionate about technology — mobile, cloud — and readily follows market trends and opportunities. He is currently focused on bringing added value to Objectivity’s retail projects.

Read More