With food waste high on the agenda for retailers under growing ESG and margin pressure, a new generation of AI-driven tools is helping them move beyond reactive markdowns towards predictive, data-led decision making, to help get ahead of the challenge, explains David Waters, chief decision scientist for software company Retail Insight.
As Earth Day 2026 focuses attention on environmental responsibility, food waste remains one of grocery retail’s most urgent challenges. Around 17% of food produced globally is wasted according to the UN Environment Programme’s Food Waste Index Report, contributing significantly to greenhouse gas emissions while placing unnecessary pressure on supply chains, margins, and resources.
For grocery retailers, this is now a strategic business priority shaped by regulatory pressure, ESG commitments and the need to operate profitably in an unpredictable market. Every item wasted has both an environmental cost and is a missed opportunity to recover margin and improve operational efficiency.
While some retailers have found significant gains from using conventional analytical solutions, waste management for many remains largely reactive and ineffective. Stores identify problems too late, often because there appears to be excess stock, and they then start price discounting before expiry dates are reached. And they are relying on manual checks and static markdown processes that struggle to keep pace with fluctuating demand. As a result, products still go unsold and waste remains stubbornly high because they are marked down too late and at an un-optimised level.
New technology provides the opportunity to reduce waste, using auto identification of item excesses. At the same time, the tech and processes need to be implemented in a way that aligns with food safety regulations, ESG commitments and store process guidelines. Retailers need to be able to experiment with AI but they must ensure that the results enable robust operations and products that enable smart pricing, underpinned by transparency, auditability, and safety.
AI-driven decision intelligence is connecting pricing, inventory, store execution and sustainability data into a single decision intelligence layer, on which retailers can move from reactive waste management to predictive waste prevention.
The foundation of this shift is visibility. Retailers need to understand which products are at risk of waste, where overstock exists (quantities excess to demand based on shelf-life constraints), how demand varies by location and time, and when intervention is required to maximise sell-through. This level of insight cannot be achieved through isolated systems or manual processes, it requires integrated, real-time data across the retail operation with full inventory granularity.
New platforms enable this by identifying waste risks earlier and providing clear, actionable recommendations. Rather than reacting to problems after they occur, retailers can anticipate them and then take proactive steps to prevent surplus stock from becoming waste. Key to this is the ability to layer excess quantity modelling of short shelf life products to maximise the universe of items that get the best price.
A central component of this approach is the move from static to dynamic markdowns.
Instead of applying fixed discounts based purely on expiry dates, machine learning or appropriate smart mathematical models analyse a wide range of variables including product type, historical sales performance, local demand patterns, seasonality and store-level inventory. And if there is a robust feedback loop, underlying forecasts improve ultimately driving greater efficiencies.
This allows retailers to determine the optimal price and timing needed to maximise sell-through where demand fluctuations or poor forecasts have created excess, while protecting margin and minimising Co2 emissions. In effect, markdowns shift from being a last-minute salvage tactic to a proactive inventory reduction strategy. The benefits are higher sell-through rates, reduced shrink, improved margin recovery and less unnecessary waste.
However, successful waste reduction cannot rely on insight alone but depends on robust store execution (i.e. compliance to discounts and timings). Even the most advanced analytics will fail if store teams are overwhelmed with tasks or lack clear direction, or at times wilfully override time and discount settings through the handset to create deeper discounts.
At Retail Insight we talk about intelligent execution where retailers deploy technologies such as prompted markdowns that enable store colleagues to receive targeted item-shelf check alerts only when intervention is required, guiding them to take the right action at the right time. This reduces manual effort while ensuring consistency and compliance across stores.
Where price discounting alone is predicted to be insufficient to exit stock through the till in customer baskets, additional exit strategies are required to avoid waste. For instance, retailers are also rethinking how they handle surplus food that cannot be sold but remains safe for consumption. Structured donation and redistribution processes are becoming an essential part of a comprehensive waste strategy.
Intelligent donation initiatives can identify suitable products, guide store teams through donation workflows, automate labelling and compliance tracking, and integrate with charity partners. This not only ensures that surplus food reaches communities instead of landfill, but also provides measurable, reportable ESG outcomes, an increasingly important requirement under regulatory and investor scrutiny.
What is becoming clear is that the future of waste reduction does not lie in isolated point solutions, it depends on integrating decision-making across pricing, inventory, store operations and sustainability initiatives. Retailers need a connected approach that allows them to act on insight quickly, consistently and at scale.
Reducing waste lowers operational costs, improves ESG performance, strengthens consumer trust and enhances in-store efficiency. At the same time, better inventory decisions improve product availability, reducing out of stocks and lost sales.
Waste should no longer be seen as an inevitable by-product of grocery retail, but as a signal of where decisions are disconnected. Retailers that unify their operations through AI-driven decision intelligence will not only reduce waste, but build more resilient, transparent and profitable businesses.
Read more from Retail Insight’s David Waters
[image credit: Retail Insight]
