Most logistics investment theses treat returns as a sub-problem of fulfillment. That framing is wrong, and it leads investors toward solutions that are too narrow to matter. Reverse logistics is not a returns management workflow. It is an entirely separate data coordination challenge — one that flows in the opposite direction of every assumption baked into the software stack built over the last twenty years.
I spent nine years in supply chain operations. I ran distribution networks where forward flow was understood, instrumented, and continuously optimized. The reverse channel was treated as an afterthought — physically, organizationally, and in terms of software budget. That experience makes me certain that the companies attacking this problem now are building something genuinely new, not just extending what already exists.
The asymmetry that makes returns hard
Forward logistics benefits from predictability. You know what you shipped, where it came from, and what condition it was in. The data model is clean: SKU, quantity, origin, destination, carrier, timestamp. The entire ecosystem of warehouse management systems, transport management systems, and carrier APIs was built around this data shape.
Returns break every assumption. The condition of a returned item is unknown until inspection. The declared reason for return correlates weakly with the actual disposition decision — a consumer marking "wrong size" on an apparel return might trigger the same processing path as a defective electronics unit, but those two items have radically different downstream options: restock, refurbish, liquidate, donate, dispose. Each option carries a different cost and recovery value, and the right choice depends on real-time information about inventory levels, liquidation market prices, refurbishment capacity, and carrier costs for secondary handling.
The forward logistics stack is not equipped to handle this decision tree. WMS systems were designed to move inventory in one direction with known attributes. Applying them to reverse flow is like using a freight manifest to run an auction — the primitives don't match.
Where the money is actually going
E-commerce return rates in European apparel run between 25% and 40% by volume depending on the category. In consumer electronics, average return rates sit closer to 10–15%, but the value-at-risk per unit is dramatically higher. A company processing 200,000 returns per month across mixed categories is looking at a coordination problem that involves dozens of disposition pathways, multiple refurbishment partners, three or four liquidation buyers, and a regulatory overlay if any items contain batteries or hazardous materials.
The cost of getting this wrong is not just the direct handling cost. It is the velocity loss — the time between a customer initiating a return and that item generating any recovery value. For a seasonal product with a 90-day selling window, a 30-day inspection and disposition cycle can eliminate the restock opportunity entirely. The item goes to liquidation at 15 cents on the dollar when it could have been restocked and resold at full margin.
This is the business case for software-native reverse logistics. Not return portal UX. Not label generation. The actual decision science about what to do with a returned item, executed fast enough that the item retains its value.
What good reverse logistics software actually looks like
The companies building here that we find interesting share a few structural characteristics. First, they orient around disposition intelligence rather than process automation. The easy version of reverse logistics software automates the return receipt and triage workflow. The hard version builds a model for what each item is worth across its possible disposition pathways at the moment it enters the return facility — and executes against that model in near-real-time.
Second, they treat condition grading as a data input, not a human judgment call. A returns facility receiving 5,000 items per day cannot rely on staff making nuanced refurbishment calls on each unit. The software needs to ingest condition signals — images, sensor data, return reason codes, purchase history — and produce a graded output that routes the item without human intervention for the majority of cases. The edge cases that require human review should be flagged, not the norm.
Third, they build toward the secondary market as a first-class output, not an afterthought. The liquidation market for returned goods is fragmented and inefficient. Buyers exist at multiple price points — recommerce platforms, B2B liquidators, regional charity networks, industrial recyclers — and pricing is opaque. A software layer that connects returned inventory to the best available buyer in real time, rather than batching everything to a single wholesale liquidator at a fixed price, can recover meaningfully more value. The business model for this layer looks more like a marketplace than a SaaS tool.
The investment landscape is catching up slowly
We backed Covariant in 2024 because their approach to warehouse robotics directly addresses the reverse logistics bottleneck — the physical handling and sorting of returned goods that currently requires manual labor. Their AI-powered robotic systems can identify, sort, and route returned items to the correct disposition path with minimal human intervention, dramatically reducing per-unit processing costs.
The broader market is still underfunded relative to the problem size. Most capital in returns technology has gone toward consumer-facing return experience — return portals, prepaid label generation, customer communication tools. These are real products solving real friction, but they don't address the core economics problem. They make the intake process smoother without changing what happens to the item after it arrives.
We're not saying return experience software is bad — it clearly has value, and there are good businesses in that space. What we are saying is that the much larger economic opportunity is in the disposition and recovery layer, and it remains significantly undercapitalized.
Regulatory pressure is about to change the calculus
The EU's Ecodesign for Sustainable Products Regulation, which became operative in 2024, changes the disposition calculus for consumer goods sold in European markets. The regulation restricts the destruction of unsold and returned goods in specific categories — textiles and electronics first, with expansion to other categories mapped out in the framework. Brands that previously disposed of returned goods as a low-cost fallback now face compliance obligations that require them to prioritize resale and refurbishment.
This regulatory shift turns reverse logistics from a cost center into a compliance function with meaningful penalties for non-performance. It accelerates the business case for software that can intelligently route returned items toward legitimate secondary channels. Companies that built this infrastructure ahead of the regulatory mandate will be positioned to onboard European retailers and brands who are now scrambling to demonstrate compliance.
The timing creates an unusual window. The problem is large, the regulation is real, and the software stack is genuinely underdeveloped. That combination — when it shows up with a technical team that has operated in the space — is exactly the kind of entry point we look for.