Returns are one of those problems that’s only going to get worse over time.
Consumer expectations have shifted permanently. Free returns, instant refunds, no-questions-asked exchanges. Amazon trained an entire generation of shoppers to treat returns as frictionless, and every other brand is now expected to match that.
There's no fixing the trend. But you can only build a better operation around it.
So every time I come across a brand that's cracked something real on returns, I like to dig in deeper.
This one's about Honeylove.
Honeylove sells shapewear. Bras, bodysuits, shorts, the kind of stuff where fit is everything and buying the wrong size is almost guaranteed if you don't know exactly what you're looking for. They've invested heavily in helping customers find the right fit upfront, with detailed sizing logic that factors in cup size, band size, and several other variables. That sophistication is also what made returns particularly brutal. Every item in their catalog has five variables: size, color, version, style, cup, and band. When a customer wanted to exchange, they had to navigate that entire matrix inside a portal built for brands with two.
Most couldn't. So they took the refund.
At hundreds of thousands of returns a year, that added up fast.
Post-purchase platform Route came in and rebuilt Honeylove’s exchange experience from the ground up. They built a custom exchange UI, embedded chat inside the portal, and integrated across four warehouse partners in four geographies.
The result: $5M in retained cash and a 20% reduction in support tickets.
I sat down with Bailey Newton, who led the project at Route, to get into exactly how they did it.
What’s Inside
Why a Standard Returns Portal Wasn't Built for a Product Like Theirs
Honeylove isn't your average apparel brand. They've built one of the most sophisticated front-end shopping experiences in intimates. Detailed sizing logic, personalized fit recommendations, a site that factors in cup size and band size to steer you toward the right product. Real engineering effort, real investment, real results.
And then a customer tries to make an exchange, and all of that sophistication evaporates.
The returns portal they were using before Route wasn't bad software. It was software built for a brand with two variables: size and color. Most brands are. Honeylove has five. So when a customer landed inside that portal looking to swap a product, they were on their own. No guidance. No logic. Just a catalog that was genuinely difficult to navigate without knowing exactly what you were looking for.
Most customers didn't know. So they didn't exchange. They refunded.
There's a thing that happens at scale where a broken experience that feels like a small friction point per customer becomes a structural leak in the business. Every single return was a moment where Honeylove could retain the sale or lose it.
And the portal, through no fault of the customer, was reliably pushing them toward the loss.
They'd already tried to fix it with their previous provider. The software wasn't the problem. The problem was the provider wasn't able to customize for their specific use case. For a brand running headless on Shopify with their own internal engineering team, that matters.
Headless means they'd gone off-road from Shopify's standard setup and built their own infrastructure on top. Any solution coming in had to be willing to work within that. Most weren't.
So the leak kept leaking.
How Route Diagnosed the Problem
Route's diagnostic process had three distinct steps. Each one built on the last, and together they gave a complete picture of what was actually broken and what fixing it would require.
Step 1: Listen to the CX Team Before Looking at Any Data
Honeylove's CX team tracks every reason a customer reaches out. When Route came in, that's where they started.
The returns process was coming up again and again. Customers said it was hard. They couldn't find the right item. They were giving up mid-exchange and taking the refund because it was easier.
That’s a specific signal: a customer who wants to exchange but can't figure out how is a completely different problem from a customer who just wants their money back. One is a lost sale you could have kept. The other is a genuine return. If you're only looking at your refund rate, you can't tell them apart.
You need to know what customers are actually saying about why they gave up.
Step 2: Benchmark the Numbers Against Similar Brands
Route has worked with over 13,000 brands. That gives them refund rates, exchange rates, and return rates across a wide range of comparable businesses. Most individual brands have no way to know whether their numbers are normal or out of line. Route does.
"We can look at a brand and say, your refund rate is higher than what we typically see for brands similar to you," Bailey said. "Here are some strategies we can actually help you with."
Honeylove's numbers were out of line. Refund rate too high. Exchange rate too low. That gap is where the money was going, and benchmarking made it visible in a way that internal data alone couldn't.
Step 3: Trace the Gap Back to Its Root Cause
With the CX signal and the benchmarks in hand, the diagnosis became clear. Five product variables, a portal built for two, and customers defaulting to the path of least resistance every time.
It was a UI problem as much as anything else. The portal wasn't surfacing the right options to the right customer at the right moment. Customers weren't choosing refunds because they wanted refunds. They were choosing refunds because finding the right exchange was too hard.
Fix the experience, fix the numbers.
What Route Needed to Build
The diagnosis pointed directly at the solution. The portal needed to actually reflect HoneyLove's catalog. That meant a custom UI built around their five variables, one that walked customers through the exchange step by step instead of dropping them into a generic selector and hoping for the best.
The Implementation
Returns don't sit in a corner of your tech stack. They sit in the middle of it. As Bailey put it: "You integrate with the e-commerce platform, but then there's usually an ERP, a WMS, a CX tool, and then bonus stuff like loyalty, tracking, and package protection."
Every one of those integrations needs to be designed, built, and tested before a real return runs through the new system.
For Honeylove, getting this right took 12 weeks.
Who Was in the Room
One thing that catches brands off guard is how many people a project like this actually involves. Returns touch everything, which means everyone has a stake in how they get solved.
On any given time throughout the implementation, Route was working with 15 to 20 people across the Honeylove business:
CTO, leading the project on Honeylove's side
Customer service team
Marketing
Supply chain and operations
If you're going into a project like this expecting a small, contained technical implementation, reset that expectation. Getting the right people aligned early, especially on the operational and CX side, is what separates a smooth rollout from one that drags.
The Custom UI
The centrepiece of the build was a custom UI inside the returns portal that walked customers through Honeylove's five product variables in a logical sequence. No more dropping a customer into a generic selector and hoping they figure it out. The portal now guides them to the right replacement.
Embedding Chat Inside the Portal
Honeylove already knew their CX team resolved tickets faster through chat than email. What they hadn't done was put chat inside the returns portal. The portal lives off the main site, so a customer who hit a problem mid-return had one move: abandon the process and reach out through a separate channel. Delays, extra tickets, and exchanges that never got completed.
Route embedded Honeylove's chat tool directly inside the portal. Customer gets stuck, customer gets help, customer completes the exchange. Nobody has to leave the flow.
Cross-Border Returns: Four Geographies, Four 3PLs
Honeylove ships worldwide, which added another layer of complexity. They don't use the same warehousing partner in every market. Their US, Canada, European, and Australian operations each have their own 3PL, and every one needed to be integrated separately.
"We had to integrate with external tools to create return labels within Canada, within Europe, and in Australia, and make sure there weren't any issues with duties," Bailey said.
That also meant building out the ability to generate return labels across different carriers in different markets and making sure duties weren't creating problems for cross-border returns. Route integrated with a global label creation tool to handle this across all of Honeylove's markets.
It's the kind of thing that doesn't come up in an initial conversation about returns software but becomes very real during implementation. For any brand operating across multiple geographies with different warehouse partners, this is a non-trivial part of the project.
The Timeline
Weeks 1 to 8: Implementation, getting every integration built and configured across the e-commerce platform, ERP, WMS, CX tools, and cross-border label creation
Weeks 8 to 12: Testing and training, making sure everything worked before real volume ran through it
The Rollout
When everything was built and tested, Honeylove didn't flip a switch and send every customer through the new system overnight.
Starting Small
The rollout started with Honeylove's customer service agents. Instead of going live on the site immediately, CX agents were given the link to the new returns portal and sent it manually to customers who reached out about a return.
At Honeylove's volume, even a small problem in the returns flow creates a lot of noise fast. Starting with agent-assisted returns meant the team could watch closely, catch issues early, and fix them before they were exposed to the full customer base. It also gave the CX team time to get comfortable with the new system before they had to support customers using it independently.
Then, Going Wide
The move to full site-wide rollout only happened once the team was confident. Not confident on paper, confident because they were already seeing results come through from the agent-assisted phase.
That's an important distinction. A lot of brands go live based on a testing checklist. Honeylove went live based on early signals that the thing was actually working. By the time the link went live on their site, there was already evidence it was doing what it was supposed to do.
Why This Approach Matters
For a brand doing hundreds of thousands of returns a year, the cost of getting this wrong at full scale is significant. A broken or confusing returns experience doesn't just create operational headaches. It creates customer service tickets, unhappy customers, and refunds that didn't need to happen.
The slow rollout de-risked the whole thing. Honeylove had a window to validate the experience before it became their primary returns channel. Any issues that came up got resolved quickly rather than sitting in a backlog.
The Results
Since switching to Route, Honeylove has:
Retained over $5 million in cash through improved exchanges
Reduced customer support tickets by 20% related to returns confusion
The $5M is simple to understand once you see it. More customers completed exchanges instead of defaulting to refunds. The product didn't change. The return volume didn't change. The portal made it easier to find the right replacement, and customers stopped giving up halfway through.
The 20% ticket reduction came from the embedded chat. Customers who got stuck mid-return previously had to abandon the flow and reach out separately. Now they get help inside the portal. Fewer abandoned returns, fewer tickets, faster resolution.
The cross-border rollout meant customers in Canada, Europe, and Australia got the same experience as US customers. Returns that were falling through the cracks in those markets started completing properly.
Read the $5M less as a one-time result and more as a rate change. Every month the exchange rate stays higher than it was before is more cash that stays in the business.
What You Can Take From This
Returns Are a Conversion Problem, Not Just a Logistics Problem
Every customer coming back to return something is still engaged with your brand. They haven't walked away yet. What happens inside that portal determines whether they leave with a refund or a new product. If your portal isn't built for your product, you are losing that conversion on every single return, and it shows up in your refund rate as if it's just normal cost of doing business.
It doesn’t have to be.
One-Size-Fits-All Return Policies Are Costing You Money
Most brands apply the same return policy to every customer. Same window, same fees, same refund timing, regardless of purchase history or return behavior.
"There's a misconception that return policy should be rigid and the same for everyone," Bailey said.
Your returns data contains very different types of customers:
Customers who buy constantly, return almost never, and keep coming back
Customers who buy once, return immediately, and disappear
Customers who have committed outright fraud
Treating all three identically subsidizes your worst customers at the expense of your best ones. Segment by return behavior and adjust accordingly:
Best customers get free returns and immediate refunds, before the item even arrives back at the warehouse
High-return customers get a return fee to deter unprofitable behavior
Fraud risks get flagged, blocked, or held for inspection before any refund goes out
Your Warehouse Integration Is Not Optional
Bailey: "If a customer returned some rocks, and you're not getting that information back from the warehouse, you can't deter them in the future."
If your returns solution isn't connected to your warehouse, you don't know what's actually coming back. You can't catch fraud. You can't make smart decisions about when to refund versus when to wait for inspection. The 3PL integration is where most of the intelligence in a returns operation actually lives.
AI Is Changing What's Possible Right Now
Route is already using AI to verify return reasons at scale. A customer claims an item is defective, they submit a photo, AI reviews it before any action is taken. On the fraud side, they're building tools to detect whether images have been manipulated or AI-generated, a growing problem as customers get more sophisticated about gaming the system.
But the bigger opportunity Bailey pointed to is cross-brand customer data. Route runs the Route consumer app, used by hundreds of thousands of shoppers across many brands. They can see purchase and return behavior across the entire network.
"We could know something about a customer from their purchase from Amazon or Walmart, look at that customer profile, and adjust the return policy even on their first purchase with you," Bailey said.
Dynamic return policies based on behavior you can see before a customer has ever returned anything to you. That's where this is going.
Get the Scope Right Before You Start
Honeylove's implementation touched their e-commerce platform, ERP, WMS, CX tools, and four different 3PLs across four geographies. It took 12 weeks and involved 15 to 20 people across every function in the company.
Returns sit in the middle of everything. Get the right people aligned before kickoff and give the project the time it actually needs.
