General
Returns Management Platforms in 2026: The Transport Questions to Ask Before You Buy
A returns management platform is software that handles the return of goods from the customer back into inventory, covering authorization, labeling, pickup or drop-off, transport, inspection and refund. Buyer evaluations concentrate on the parts that demo well: the customer-facing portal, the label logic, the refund rules and the disposition engine. The transport layer, which decides when a pickup actually happens and what it costs, is usually assessed on whether it exists rather than on how it behaves. That is the layer where the economics live, because a single blended cost-per-pickup figure conceals a range of more than three times across zones inside the same network. Locus, the world’s first Decision-Intelligent, Agentic TMS, sits in that layer, solving pickup scheduling and reverse routing against more than 250 real-world operating constraints rather than treating them as a downstream handoff.
Key Takeaways
- A dedicated pickup route needs a minimum number of stops in a zone on a day to be economic. Below that, the pickup is standalone and costs roughly 3.5 times as much.
- At 2.5 returns per zone per day, 96% of zone-days fall below viable route size. At 20 per day, none do.
- A network-blended $3.33 per pickup hides a spread from $1.50 in dense urban zones to $5.20 in rural ones, describing no zone that exists.
- The test that separates platforms is whether the system can decline a pickup window it cannot serve economically, rather than accepting every booking the calendar allows.
- First-attempt rate multiplies with density. At 78% first-attempt success a perfectly dense zone still pays $2.97 per completed pickup against $1.50.
- Locus computes pickup feasibility against live network density, so a program can see which zones subsidize which instead of managing to an average.
Why the Transport Layer Decides Returns Cost: The Business Case
The volume makes small per-unit differences material. The National Retail Federation’s 2025 Retail Returns Landscape, produced with Happy Returns, puts returns at 15.8% of annual retail sales in 2025, worth $849.9 billion, with online returns running at 19.3%. At that scale a dollar of difference in cost per pickup is a budget line, not a rounding error.
Returns are also now a purchase criterion rather than an after-sales process. The same research found 82% of consumers cite free returns as a major consideration when buying, up from 76% the previous year. A retailer offering free returns has absorbed the transport cost into margin, which makes the efficiency of that transport a pricing question.
The cost range in the reverse leg is wide and published. Locus data puts standalone return pickups at $4 to $6 per item against under $1.50 when the pickup is consolidated into an existing route. That gap is not a software feature, it is a density outcome, and density is a property of the network rather than of the platform.
The carrier options underneath all this have widened. The Pitney Bowes Parcel Shipping Index records non-major carriers doubling their share of US parcel revenue from 3.4% to 7.2% in a 23.1 billion parcel market, which matters for returns because regional operators and consolidation networks are exactly what make thin-zone collection viable. Meanwhile the cost of putting a vehicle on the road keeps climbing: ATRI’s 2026 analysis puts the industry-average operating cost at $2.336 per mile in 2025, the highest in the report’s history. A trip made for one collection is more expensive than it was, in a leg that was already the hardest to fill.
The forward leg sets the benchmark for what good looks like. McKinsey finds that raising drops per stop from one to five cuts labor and vehicle cost by more than 50%. The same arithmetic governs pickups, which is why a returns management platform that cannot influence stop density cannot influence cost.
What the Transport Layer Actually Decides
The model below assumes a dedicated pickup route needs six or more stops in a zone on a given day to be economic, with returns arriving randomly across zones. Rates and thresholds are illustrative; substitute your own and the structure holds.
1 The window is offered before anyone knows the zone will be viable
A customer requests a return and the platform offers pickup windows. In most systems those windows come from a calendar, constrained by operating hours and lead time, not by whether that zone will have enough volume that day to justify sending a vehicle. The promise is made first and the economics are discovered afterward.
That is defensible when the platform has no visibility of network density, which is the normal case when returns software and transportation run as separate systems. It stops being defensible when the same organization owns both, because the information needed to make a better promise exists and is not being used.
2 Below a density threshold, every pickup is a standalone trip
| Returns per zone per day | Zone-days below viable route size | Blended cost per pickup | Against dense zones |
|---|---|---|---|
| 1 | 100% | $5.20 | 247% more |
| 2 | 98% | $5.14 | 243% more |
| 3 | 92% | $4.89 | 226% more |
| 5 | 62% | $3.78 | 152% more |
| 8 | 19% | $2.21 | 47% more |
| 12 | 2% | $1.58 | 5% more |
| 20 | 0% | $1.50 |
The curve is steep between three and twelve returns per zone per day, which is exactly the range most suburban and secondary markets occupy. A platform that reports one cost-per-pickup number for the network is averaging across both ends of this curve.
3 One network contains several of these at once
| Zone type | Returns per zone per day | Share of network | Zone-days below viable | Cost per pickup |
|---|---|---|---|---|
| Dense urban | 20 | 30% | 0% | $1.50 |
| Suburban | 6 | 35% | 45% | $3.15 |
| Small town | 2.5 | 25% | 96% | $5.04 |
| Rural | 0.8 | 10% | 100% | $5.20 |
| Network blended | $3.33 |
The blended figure of $3.33 is arithmetically correct and operationally useless. It describes no zone in the network. A returns program managed to that number will over-invest in dense zones where there is nothing left to win and under-invest in the thin zones carrying the cost.
This is the first question to put to a vendor: not what the average cost per pickup is, but whether the platform can produce this table for your own network.
| Also Read: US Returns Hit $850 Billion in 2025: Why US Retailers Are Restructuring Reverse Logistics in 2026 |
|---|
4 The decisive capability is the ability to say no
Everything above points at one test. A returns management platform that accepts every booking the calendar permits will generate uneconomic pickups in thin zones indefinitely, because nothing in the flow is empowered to refuse. A platform connected to transportation can do one of three things instead: offer a later window when the zone will have density, offer a drop-off alternative, or accept the booking and price it correctly so the cost lands where the decision was made.
Ask to see the system decline or reprice a window in a demonstration. Most cannot, and the inability is structural rather than a configuration gap.
5 Drop-off is a routing decision, not a customer preference
Consolidation networks and carrier drop-off points exist precisely to solve the thin-zone problem, by moving the density problem from the pickup to the drop-off location. Whether a given return should be collected or dropped off depends on that customer’s zone density that week, which is a routing calculation.
Most platforms present drop-off and pickup as a customer choice with no cost signal attached. That is a reasonable experience decision and an expensive default, because the customer has no information about which option the network can serve cheaply and no reason to care.
6 Carrier assignment for reverse flows is a different problem from forward
A forward delivery has a known origin, a fixed departure time and a full vehicle. A pickup has a customer-controlled availability window, an unknown item size until collection, and a failure mode where nobody is home and the item is still with the customer. Carrier scorecards built on forward performance do not predict reverse performance, and rate cards built on forward volume rarely price the difference.
That failure mode has a price, and it compounds with density rather than sitting alongside it. A re-attempt has no other stops behind it, so it is planned and paid for as a standalone trip regardless of how dense the zone was on the first pass.
| First-attempt success | Attempts per completed pickup | Effective cost in a dense zone | Against a single attempt |
|---|---|---|---|
| 95% | 1.05 | $1.77 | 18% more |
| 90% | 1.11 | $2.08 | 39% more |
| 85% | 1.18 | $2.42 | 61% more |
| 78% | 1.28 | $2.97 | 98% more |
| 70% | 1.43 | $3.73 | 149% more |
| 60% | 1.67 | $4.97 | 231% more |
At a 78% first-attempt rate, a zone with perfect density pays $2.97 per completed pickup rather than $1.50. An operation can therefore hold best-in-class consolidation and still land at suburban economics, purely on attempt rate. The two levers multiply, and most returns programs measure neither.
Ask how the platform assigns carriers to pickups specifically, whether reverse performance is scored separately, and what the first-attempt rate is by carrier. If the answer is that the same allocation logic applies to both directions, the platform is treating a pickup as a delivery run backwards, which it is not.
7 What to run before the evaluation starts
The analysis that makes this concrete uses data a retailer already holds. Take a quarter of completed returns with customer postcode, request timestamp and actual pickup cost. Group by zone, count returns per zone per day, and plot cost per pickup against that count. The threshold where cost falls sharply is your viable route size, and the share of your volume sitting below it is the part of the program no platform will fix by itself.
Bring that to the vendor conversation. It converts a feature discussion into a capacity discussion, and it is the fastest way to find out whether a returns management platform has a transport layer or a transport integration. Run the same grouping for first-attempt rate, because the two drivers interact and a zone can be failing on either.
The output is a short list of zones where the returns program is structurally expensive, and that list is more useful than any vendor comparison. It tells you whether your problem is density, in which case consolidation and drop-off are the answers, or attempt rate, in which case the answer is scheduling and communication rather than routing. Those lead to different platforms, and a buyer who has not separated them will be sold the wrong one convincingly.
Returns Platform vs Transport-Native Returns: Key Differences
| Dimension | Point-solution returns platform | Returns handled in the TMS |
|---|---|---|
| Pickup window offered from | Calendar and lead time | Network density and capacity |
| Cost visibility | Blended average | By zone and by day |
| Response to a thin zone | Accepts the booking | Reprices, reschedules or offers drop-off |
| Carrier assignment | Forward scorecard | Reverse performance scored separately |
| Consolidation with deliveries | Not available | Native, same fleet and plan |
| Where the cost surfaces | Monthly carrier invoice | At the point of booking |
What to Look for in a Returns Management Platform
Cost per pickup by zone, not blended. Ask for the table in step three against your own postcodes. A platform that can only produce a network average is not modeling the thing that drives your cost.
The ability to decline or reprice a window. This is the single clearest separator. Ask to see a booking refused or repriced in a demonstration, and ask what rule produced it.
Reverse-specific carrier scoring. Pickup success, first-attempt rate and dwell on collection behave differently from delivery metrics. Confirm they are measured separately rather than inherited from the forward scorecard.
Drop-off as a routed option. The system should be able to recommend drop-off where pickup is uneconomic, with the reasoning visible, rather than presenting both as equivalent customer choices.
Consolidation with forward routes as a native capability. If pickups can only be planned as their own routes, the cheapest option in the network is unavailable by design. Ask how a pickup enters an existing delivery route and what lead time that requires.
First-attempt rate reported by carrier and by zone. Attempt rate and density multiply, so a platform that reports neither cannot explain its own cost. Ask which of the two is driving your most expensive zones, and expect the answer to differ between them.
Returns Transport in Action: Real-World Results
A Canadian grocery brand delivering fresh and perishable orders across more than 30 cities ran its last mile through contracted third-party fleets, which meant both delivery and collection depended on partners it did not directly control. In the carrier orchestration deployment, moving allocation and tracking into a single orchestration layer produced 33% faster deliveries, 15% lower fulfillment cost, 25% less manual shipping time and customer support resolution 10 to 20 times faster. The support number is the relevant one for returns, because collection exceptions generate contacts at a far higher rate than deliveries do.
A leading North American retailer with several hundred stores consolidated ocean, rail and road execution off six legacy systems in a multimodal automation program. Exceptions resolved in under two hours against on-time store delivery above 99%, with route compliance above 95% and manual dispatch reduced more than 80%, returning more than $1M in savings and breaking even in year one. Store networks are also the consolidation points most returns programs depend on, which is why the same platform decision governs both flows.
Common Returns Management Evaluation Mistakes to Avoid
Evaluating on the customer portal. The portal is the cheapest part of the system to build and the easiest to demonstrate. The transport layer is where the cost variance lives and it demonstrates poorly.
Accepting a blended cost per pickup. One number across a network containing both 20-returns-a-day zones and 2.5-returns-a-day zones describes neither. Ask for the distribution.
Assuming reverse carrier performance follows forward performance. A carrier that delivers well may collect badly, because the failure modes differ. Score them separately before committing volume.
Treating drop-off as purely a customer preference. Where pickup is uneconomic, drop-off is the routing answer. Presenting the two as equivalent choices with no cost signal guarantees the expensive one gets chosen often.
Improving density and ignoring attempt rate. The two multiply, so a consolidation program that lifts a zone to twenty returns a day while first-attempt success sits at 78% delivers roughly half the saving the business case assumed.
How Locus Handles the Returns Transport Layer
Locus, the world’s first Decision-Intelligent, Agentic TMS, treats a pickup as a routing decision made against live network state rather than a booking accepted from a calendar. The route planning system evaluates whether a zone will carry enough density on a given day to serve a collection economically, so the window offered reflects what the network can actually do, and a pickup that belongs in an existing delivery route is placed there rather than planned as its own trip. Because the Capacity, Dispatch and Carrier agents hold forward and reverse flows in the same plan, reverse carrier performance is scored on its own terms and the cost of each pickup is visible at the point of booking rather than on a carrier invoice weeks later.
Locus is recognized by Gartner for seven consecutive years, featured in the 2026 Hype Cycle for Supply Chain Execution and Logistics Technologies, named a Leader in TMS by QKS Group in the SPARK Matrix, and ranked #1 in Route Planning on G2’s 2026 Best Software Awards. In October 2025, Ingka Investments, the investment arm of Ingka Group, the world’s largest IKEA retailer, acquired Locus. Locus continues to operate independently.
A returns management platform is usually chosen on the parts that demonstrate well and lived with on the part that does not. The transport layer decides cost, and it decides it by zone density rather than by feature: below the threshold where a dedicated route becomes economic, every pickup is a standalone trip at roughly 3.5 times the consolidated cost, and at 2.5 returns per zone per day that describes 96% of zone-days. A blended network figure of $3.33 conceals a spread from $1.50 to $5.20 and describes no zone that actually exists. The test worth running is whether the platform can decline or reprice a window it cannot serve. Locus makes that decision against live density in the same plan as the forward network. Request a Locus returns transport assessment to see the distribution behind your own average.
Frequently Asked Questions
What is a returns management platform?
A returns management platform handles the movement of goods from the customer back into inventory, covering return authorization, labeling, collection or drop-off, transport, inspection and refund. Platforms differ most in how much of the transport leg they actually control, which is also where most of the cost variance sits.
What should I ask a returns management vendor about transport?
Ask for cost per pickup by zone rather than blended, ask to see the system decline or reprice a pickup window it cannot serve economically, and ask how a pickup gets added to an existing delivery route. Those three answers separate a transport layer from a transport integration.
Why does cost per pickup vary so much across a network?
Because a dedicated pickup route needs a minimum number of stops in a zone on a day to be economic. Above that threshold, collections consolidate and cost falls toward $1.50 in this model. Below it, each pickup is effectively a standalone trip at around $5.20, and at 2.5 returns per zone per day, 96% of zone-days fall below the threshold.
Is a blended cost-per-pickup figure useful?
Not for managing the program. In the modeled network, a blended $3.33 spans zones ranging from $1.50 to $5.20 and describes none of them. It hides which parts of the network subsidize which, so improvement effort gets directed at dense zones where there is little left to gain.
Should returns be collected or dropped off?
It depends on zone density, which makes it a routing decision rather than a customer preference. Where a zone will not reach viable route size, drop-off at a consolidation point or carrier location is usually the cheaper option, and a platform connected to transportation can recommend it with the reasoning visible.
Do reverse and forward carrier performance track together?
Not reliably. A pickup has a customer-controlled availability window, unknown item dimensions until collection, and a distinct failure mode where nobody is home and the goods remain with the customer. Score reverse performance separately before allocating volume on the basis of delivery results.
General
ELD Revocation and Fleet Utilization in 2026: What Happens to Capacity When Your Hours-of-Service Feed Goes to Paper
Sep 16, 2026
FMCSA has pulled 79 devices from the ELD registry since January 2025. If your dispatch uses hours-of-service as a live input, a revocation costs capacity utilization, not just compliance.
Read more
General
Returns Management in 2026: Why You Cannot Consolidate a Pickup You Do Not Know About Yet
Sep 16, 2026
Consolidating return pickups cuts cost by roughly 70%, but only for returns the planner knows about before it commits the route. At an 18-hour commit, 45% are invisible.
Read moreInsights Worth Your Time
Returns Management Platforms in 2026: The Transport Questions to Ask Before You Buy