Treat a diversion program as an information system as much as a logistics one. The organizations moving the needle on how much food reaches a landfill are the ones treating their programs as an intelligence source.
By Ben Kuethe Oaks
Across the food waste diversion industry, the conversation has shifted. It is no longer just about whether organizations should be diverting food from landfill, it is about how well their programs are actually built to do it. Many of these programs were not designed so much as assembled piece by piece: a hauler for one geography, a composter for another, and a landfill contract as a failsafe when the patchwork ultimately breaks down. Collection volumes keep climbing and processing capacity keeps expanding, but volume diverted and waste reduced are not the same metric. Too many programs, whether run by a generator, a hauler, a processor, or a municipality, are optimized for the first without ever touching the second.
The scale of what moves through these programs is significant. In grocery retail alone, perishable shrink, product that never reaches a customer, accounts for 65 percent of total store shrink, while representing only 38 percent of sales. The average grocery store loses roughly $82,000 a year to produce shrink alone, another $93,000 to meat, and $74,000 to deli. Across a 300-store regional chain, that is $75 million a year in three departments. Grocery is not unique. Multiply that pattern across manufacturing, foodservice, and distribution, and it becomes clear why the industry’s approach to diversion must be judged by more than tonnage collected.
Whether you are building the infrastructure that processes this material, managing a municipal collection program, or advising generators on how to structure theirs, the following five practices are worth considering. What they require is treating a diversion program as a system built to reduce waste, not just to move it efficiently.
Photo courtesy of Divert.
#1: Evaluate Programs By What They Prevent, Not Just What They Remove
A collection or processing service that reliably moves material out is solving yesterday’s problem, but a program built to reduce waste asks a different question up front: what generated this material and how do we address that source? That requires a feedback loop that connects data on what is going unsold or unused back to purchasing, ordering, and production decisions further upstream. Removal and reduction are not the same service, even when they are sold or reported as one. The test for any program is whether it is designed to reduce volume over time or just to manage what shows up on pickup day.
#2: Consolidate the Chain to Reduce Coordination Headaches and Points of Failure
A patchwork of collection, processing, and disposal relationships is common across the industry, whether it was assembled by a generator juggling multiple contracts or a municipality coordinating separate hauler and processor agreements. It introduces the same structural risk regardless of who built it. Coverage gaps, capacity constraints, and unplanned costs can arise whenever any single link fails. Consolidating around fewer partners does not just reduce administrative overhead, it also frees up the operational bandwidth that would otherwise go toward managing the patchwork—bandwidth that could instead go towards answering the harder question of why the volume exists in the first place.
#3: Know Where the Material Actually Ends Up and How it is Managed
Brokered relationships, where a hauler or waste management company arranges for material to go somewhere without directly controlling the infrastructure, are common throughout the industry. This leads to a loss of customer visibility into how their material is managed. If, for example, a downstream processor mixes shredded packaging with organics and then sends that material to animal feed or compost, that introduces a new layer of risk for the customer who generated the material. The risk is also structural: when a broker’s downstream processor, composter, or digester hits capacity, the material must go somewhere else, and somewhere else is often a landfill. This is the same dynamic municipalities run into when trying to meet organics diversion mandates through contracted haulers or processors. If the downstream facility is capacity constrained, diverted material can end up in a landfill anyway, undermining the mandate it was designed to satisfy, often without anyone finding out until after it has happened. Knowing who the actual processor of record is, and what the contingency is if that processor’s capacity is exceeded, is basic due diligence for any organization relying on someone else’s processing capacity.
#4: Make Data Capture Part of the Operation, Not An Extra Task for Already-Stretched Staff
Understanding what is in a food waste stream—composition, volume, source, seasonality—is the foundation of any real improvement effort, and it is also the piece most programs get wrong. In theory, frontline staff can log what goes unsold as part of their normal process. In practice, the person handling near-date product is also receiving deliveries, stocking shelves, and helping customers, and data entry is the task that gets skipped first. The result, industry-wide, is fragmented records that cannot be aggregated across sites or connected to what happens further upstream. Programs that capture data as a byproduct of the diversion process itself, at the point of pickup or processing, where consistency can be managed and enforced, produce a fundamentally more reliable picture of the stream.
#5: Treat the Data as an Input to Upstream Decisions, not Just a Report Card
Knowing how much material moved through a program last month is a start. The more useful question is what that data reveals about why. Connected to ordering history, production schedules, or collection patterns, waste stream data starts to surface root causes: a chronic over-buy in a specific category at one site, a markdown failure that repeats every week, a product mix that consistently does not match what a market needs. These are fixable problems, but only if the data is structured well enough to surface them, and only if someone is positioned to act on what it shows.
An Information System
What these practices require, more than any specific technology or vendor, is treating a diversion program as an information system as much as a logistics one. A fragmented, backend-only approach may look efficient measured one contract at a time, but it rarely generates the visibility needed to change what happens upstream, at a generator, or inside a processing facility’s own operations. The organizations moving the needle on how much food reaches a landfill are the ones treating their programs as an intelligence source and using what that intelligence tells them to prevent the next load before it is ever generated. | WA
Ben Kuethe Oaks is SVP of Commercial at Divert, Inc., a vertically integrated circularity platform operating inside major U.S. grocery retailers and CPG companies. For more information, visit .