The Industry Got the Measurement Right. Now Comes the Harder Question.

When the grocery business is examined closely, one of the more striking facts is not the scale of the waste. It is how thoroughly everyone already understands it. Ask a store manager about near-date inventory in the perishables section and they know exactly which products are three days from expiration, which categories run the highest shrink, and roughly what it costs the store each week. The information exists. What is missing in most operations is a system that acts on it automatically, before the product becomes a loss rather than a sale.

That distance between knowing and acting is the lens through which the U.S. Food Waste Pact’s second annual report deserves to be read. The headline numbers are real progress: unsold food rates in retail fell 1.1 percent year-over-year, the foodservice efficiency metric improved by 5.7 percent, and pilot programs averaged more than 50 percent waste reduction across participating companies. For a problem the size that ReFED documents in its annual analysis, 73.9 million tons of surplus food generated in the United States in 2023 alone and representing 31 percent of the entire food supply, this is movement in the right direction. It is also a description of a gap the report itself doesn’t quite name.

What the Denominator Change Actually Tells Us

Before examining that gap, the measurement work deserves credit, because it is not trivial. For most of the past decade, food waste reporting in retail was effectively incomparable across companies. Different organizations tracked different things at different points in the supply chain, using definitions that couldn’t be benchmarked against each other. Raw tonnage was the most common external metric, and it systematically penalized high-volume operators while making trend analysis nearly meaningless.

The shift to rate-based metrics, unsold food as a share of total inventory moved in retail and waste per unit handled in foodservice, solves a real structural problem. You cannot manage what you cannot compare, and for years the industry couldn’t compare anything that held up to scrutiny. McKinsey’s research into food loss across the farm-to-retailer supply chain makes clear that creating transparency and setting shared targets are the foundational prerequisites for action, before any technology or operational initiative can take hold. The pact’s measurement work is that foundation.

The 2024 data makes the value of this concrete: more food moved through the retail signatories’ systems last year, yet the waste rate still fell. Without rate normalization, that result would have looked like backsliding. The methodology revealed the progress that the raw numbers obscured, and the pact deserves recognition for building that capability before declaring results.

Why the Pilot Numbers Deserve Scrutiny

The pilot results are where friction is warranted, because 50 percent average waste reduction is a number that can do a lot of work in a press release and much less work in a boardroom conversation about scale.

Pilots succeed under conditions that don’t replicate automatically. They attract the most motivated store managers, dedicated project resources, close monitoring, and organizational attention from senior leadership. They run in environments where participating teams have explicitly committed to operating differently for the duration of the project. Those conditions are not what exists in a 900-store chain where replenishment decisions are embedded in procurement systems built fifteen years ago, where category managers are evaluated on in-stock rates and promotional execution, and where waste efficiency has no dedicated line in anyone’s performance review.

Academic research on perishable pricing supports the underlying logic clearly enough. Studies in sustainability and operations management journals have consistently found that consumers’ willingness to pay for perishables decreases as products approach their sell-by date, and that retailers who reduce prices systematically in proportion to remaining shelf life see both reduced waste and measurably stronger customer trust and return intent. The pricing logic is sound and the peer-reviewed case for acting on it is established. What keeps pilot results from traveling is not analytical uncertainty. It is organizational resistance, and that is a different problem entirely.

The Incentive Architecture Is the Real Constraint

Here is the structural problem that better measurement, on its own, does not fix: the people making daily decisions about ordering, markdowns, and inventory levels in most retail organizations are not measured on waste. They are measured on in-stock rates, gross margin, and promotional compliance. Waste, where it appears at all in their performance framework, is typically buried in a shrink line that aggregates theft, damage, and expiration without distinguishing among them. A manager who cuts waste by 30 percent while seeing a 2 percent dip in in-stock performance will not, in most organizations, be viewed as having had a good quarter.

This misalignment has a consumer cost that is increasingly well-documented. Deloitte’s research on fresh food consumer preferences found that 57 percent of consumers prefer to shop at retailers they perceive as meaningfully reducing food waste. Most internal performance frameworks give operators no mechanism to respond to that preference. The business case is visible in the consumer data. The incentive to capture it is absent from the performance review.

Until waste efficiency sits alongside gross margin as a metric that determines how a category manager or store director is evaluated, the gap between pilot performance and operational reality will persist. The pact’s reporting framework creates external accountability. The internal incentive architecture is a separate problem, and it belongs to each company’s leadership to solve.

The Transparency Dimension the Industry Keeps Skipping

There is an angle in the food waste conversation that is consistently underweighted, and it has nothing to do with supply chain optimization or forecasting algorithms. It is what the customer actually experiences.

In most retail environments, near-date inventory is managed entirely out of the customer’s view. The markdown decision, when it happens at all, is made by a department manager walking the floor. The customer has no systematic way to know the freshness status of what they are considering buying. That information gap is not neutral. Near-date inventory ends up with no legitimate price signal in the market: it can be discarded, donated, or occasionally caught by an attentive shopper who checks the label, but it cannot be sold at a price that reflects its actual remaining value because no mechanism exists to connect remaining shelf life to the customer’s purchase decision.

Deloitte’s fresh food research found that 77 percent of consumers value fresh food partly for its role in preserving freshness and reducing waste, and that more than half use label data to determine which fresh product to purchase. The appetite for freshness transparency is documented and substantial. What is missing in most retail environments is an operational system that converts that appetite into a completed transaction.

Research on perishable retailing has found that matching price to remaining freshness transparently strengthens consumer trust and increases revisit intention. The near-date product stops being a liability and becomes a value proposition for price-sensitive shoppers. The retailer recovers margin on inventory that would otherwise generate zero revenue. The e-commerce channel makes the opportunity particularly visible, because the product detail page is a natural place to surface freshness data. In physical retail it requires consistent labeling, real-time pricing updates, and a merchandising logic that treats near-date inventory as a promotional category rather than a disposal problem. Neither is technically complex. Both require organizational intention that most chains have not yet applied.

What Precompetitive Collaboration Can and Cannot Do

The pact’s precompetitive architecture is the right model for the measurement problem. Sharing waste rate data and process insights across competitors does not compromise anyone’s ability to execute better than their peers. It raises the floor for the whole industry, which benefits retailers, consumers, and the broader food system. The initiative doubled its signatory base in 2025, and the methodology it has established, normalized rate metrics, year-over-year comparability, documented pilot results, is a real contribution to the industry’s capacity to act on what it already knows.

But precompetitive collaboration has structural limits worth naming plainly. It can establish shared metrics. It can document what works in pilots. Compelling companies to redesign their internal performance frameworks is beyond its scope, as is mandating the integration investment that connecting inventory data to pricing logic requires. Voluntary frameworks cannot overcome the organizational inertia that keeps proven approaches trapped in proof-of-concept status long after the concept has been thoroughly proved.

Research published by the Ellen MacArthur Foundation and Google, with analytical support from McKinsey, estimated that AI applied to designing out avoidable food waste could generate an economic opportunity of up to $127 billion annually by 2030. The analytical case for investment is not contested. What remains unsettled at the operational level is whose budget pays for the change and whose workflows have to adapt. Those are organizational and political questions, not technical ones, and no external reporting framework settles them.

The Number That Should Be More Uncomfortable

The 2024 pact data establishes a real baseline and documents real progress. The unsold food rate in retail fell 1.1 percent. The foodservice efficiency rate fell 5.7 percent. Controlled pilots averaged 50 percent waste reduction. Those are results the participating organizations earned.

They need to be read against a harder backdrop. ReFED’s 2025 annual report is unambiguous on the national picture: surplus food in the United States has rebounded to roughly 2016 levels, the baseline year for the national goal of halving food waste by 2030, after a temporary decline during the pandemic. In ReFED’s own assessment, the country is “still far off from meeting the 2030 reduction goal,” and it is “becoming increasingly difficult to see how we will reach it.” The pact’s pilot ceiling and the national operational baseline are not in the same conversation yet.

The industry should be more troubled by the distance between what the pilots show is achievable and what the national baseline shows is actually happening than it is encouraged by the headline figures. The pilots capture what is possible when organizations align around a specific goal with full resources and attention. The national data captures what happens under normal conditions: competing priorities, legacy systems, and incentive structures that were never designed with waste reduction in mind.

Closing that distance requires changes to how performance is defined inside organizations, investment in the infrastructure that connects inventory data to pricing decisions in real time, and a willingness to make the customer a participant in freshness management rather than a bystander. The consumer data shows customers are ready to respond. The organizations serving them are not yet ready to ask.

View Profile

Lucas Piccinin

Lucas Piccinin is the Founding Principal of NetConv, a retail technology company that helps retailers worldwide optimize the pricing and placement of perishable goods using real-time expiration data.

Through NetConv, Lucas works with retailers to bring supply and demand for perishables into equilibrium, reducing food waste and increasing profitability. The company’s intelligent pricing system applies gradual discounts and strategic repositioning, ensuring products sell through before they expire.