Latest / The DTC Marketing Podcast with Fexingo: Direct-to-Consumer Brands, Shopify, and E-Commerce Growth / How DTC Brands Use Warehouse Robots to Cut Fulfillment Costs
Transcript
- Lucas: Okay, so there's this stat that I think really frames the problem for a lot of DTC brands right now. According to a 2025 logistics survey from Ware2Go, nearly 60 percent of DTC companies say the single biggest pressure on their margin is fulfillment cost — not raw materials, not ad spend, but getting the box out the door. Luna: That tracks with what I've been hearing. Shipping rates keep climbing, plus labor is harder to find and more expensive. So where do brands find the lever? Lucas: Right — and the lever a surprising number of them are pulling is warehouse robotics. Not the giant Amazon-style arm systems, but autonomous mobile robots — basically rolling shelves that bring product to the picker instead of the picker walking to the product. And one case I think is really instructive is Bear Mattress. Luna: Bear — the mattress-in-a-box brand that's owned by Propel? I think I've seen their ads. Lucas: That's the one. So Bear operates a roughly 100,000-square-foot fulfillment center in New Jersey. Before robots, their pickers walked an average of 12 miles a day per shift. Twelve miles. That's basically walking a half-marathon every shift just to pick orders. They deployed a fleet of 40 Locus Robotics bots, and instantly that walking dropped to under two miles per shift. Luna: Forty bots — what did that cost them, roughly? Lucas: Locus operates on a robots as a service model, so Bear pays per bot per hour. Industry estimates put it around eight to ten dollars per bot per hour depending on volume. For 40 bots running two shifts, that's somewhere in the ballpark of six to eight thousand dollars a week. But the flip side — Bear told me their labor costs on the picking line dropped by about 30 percent, and they increased same-day order capacity from 5 percent of orders to 40 percent. So the return on investment came in under 18 months. Luna: Eighteen months is pretty quick for a capital decision. And I'm guessing the robots also let them scale without hiring more people during peak seasons. Lucas: Exactly. That's the hidden ROI. Most DTC brands have massive seasonal swings — Black Friday, Cyber Monday, holiday returns. With manual picking, you either overstaff the rest of the year or you scramble to hire temps. The robots let you flex capacity without a hiring spike. Bear said they handled 140 percent more volume during peak 2025 without adding a single temporary picker. Luna: So the math seems pretty clear for a brand doing Bear's volume. But what about a smaller DTC brand — maybe doing 100 orders a day instead of a thousand? Can they make the numbers work? Lucas: Great question. The answer is yes, but the threshold matters. The smallest viable deployment I've seen is from a company called 6 River Systems — they're owned by Shopify now, actually — and their bots start at around 10 to 15 units. For a brand doing 200 orders a day, a 10-bot deployment can cost roughly $2,500 a week. The labor savings if you replace two pickers at, say, $18 an hour — that's about $1,440 a week per picker. So you need to replace at least two pickers to break even. At 200 orders a day, you probably already have three or four pickers, so the math can work. Luna: But that assumes you can actually reduce headcount. Some brands might not want to lay people off. They're more likely to reassign workers to quality checks or customer service. Lucas: That's actually what most of them do. The pitch from the robotics vendors isn't 'fire your team.' It's 'stop hiring more.' Especially in a tight labor market. And the workers themselves prefer it — pacing around concrete floors all day is physically brutal. Turnover in manual picking can be 100 percent a year. With robots, turnover drops. Workers feel like they're operating technology instead of being a cog. Luna: Yeah, I can see that. There's a dignity angle too. I'd rather manage a fleet of robots than walk 12 miles carrying boxes. Lucas: Exactly. And the tech keeps getting cheaper. Locus recently introduced a lighter bot called the Vector that costs about 30 percent less per hour than their standard model. And there are even newer entrants like RightHand Robotics that combine a robot arm with a picking gripper — no human involved at all for certain items. That's still expensive — we're talking six figures per unit — but it's coming down. Luna: I want to pause on something you said earlier — about Bear increasing same-day capacity. That's not just a cost play, that's a customer experience play. Same-day shipping is becoming table stakes for DTC, especially in categories like mattresses where customers are comparison shopping. Lucas: Huge point. And this is where the marketing angle connects directly. If you can offer 'order by 2 PM, delivered by 8 PM' while your competitor says 'ships in 3–5 days,' you win the conversion. Bear invested in robots partly because they saw their cart abandonment rate spike when delivery estimates exceeded three days. They ran an A/B test — showing a same-day badge versus a 2–3 day badge — and the same-day badge lifted conversion by 18 percent. Luna: That's a clean experiment. And it makes the robotics investment a marketing decision, not just an operations one. Lucas: Right. You can't run a marketing campaign around two-day shipping if your warehouse can't actually deliver it. The robots let the brand promise something real. And then they advertise that promise. Luna: I think there's also a second-order effect on returns. Faster delivery — especially for something like a mattress — means the customer gets it sooner, tries it sooner, and if they don't like it, they return it within the trial window. But since the warehouse picks and packs faster, the return processing can also be faster. That's a whole cycle. Lucas: You're ahead of me — Bear actually reorganized their returns processing onto the same robots. Inbound returns get scanned, sorted by condition, and routed to a restock zone or a discount zone — all on the same bot network. They told me they cut return processing time from 48 hours to 12. That improves cash flow because returned inventory goes back onto the shelf faster. Luna: That's smart. So the robot fleet becomes a platform for both outbound and inbound logistics. Lucas: Now, a lot of DTC brands are asking whether they should buy their own bots or go the robots as a service route. I think for most, the RaaS model wins because it's an operating expense, not a capital expense. You don't need to raise a round or take a loan. You just pay a monthly fee that scales with usage. And if the brand grows, you add more bots. If demand dips, you return them. Luna: That flexibility is especially important for smaller brands that don't have deep pockets. But what about brands that are really small — like 50 orders a day? Is there any option? Lucas: At that volume, the math gets harder. But there are micro-fulfillment solutions — companies like Fabric or Takeoff Technologies — that operate automated micro-hubs. You essentially rent a room in their facility. They pick and pack your orders using their robots, and you pay per order. Usually around $3 to $5 per order, which is competitive with a manual picker when you factor in the walking time. So even a tiny brand can access robot efficiency without deploying anything on-site. Luna: That's basically fulfillment as a service — white-label robotics. I could see that becoming the default for early-stage DTC brands. Lucas: It's already happening. And the interesting thing is that these micro-hubs are located in dense urban areas — not in the middle of nowhere. So same-day delivery becomes possible in cities, which is where most DTC customers live. That's a marketing story in itself. Luna: Let's zoom out. We've been talking about hardware and per-order costs, but I wonder about the software layer. How does the AI optimization actually work? Because it's not just 'robot goes here, picks item, goes there.' There's a routing algorithm. Lucas: Absolutely. The brains of the system — at least for Locus — is what they call the LocusHub. It's a cloud-based AI that coordinates the fleet. Every time an order comes in, the system decides which bot should go to which shelf in which order to minimize total travel time. It also learns over time which products are picked together most often and pre-stages them closer to the packing station. Bear said the AI cut their average pick path by another 15 percent after the first three months of operation. Luna: So the system gets smarter the longer it runs. That means the ROI improves over time, not just from labor savings but from increasing throughput. Lucas: Exactly. And there's a machine learning layer that predicts order volume by hour of day. So the system can pre-position bots near high-demand pick zones before the orders even come in. It's anticipatory logistics. Luna: That's wild. So you could have a scenario where the bots are already rolling toward the mattress aisle at 9 AM because the AI knows that's when the biggest batch of orders from the overnight period will be released. Lucas: That's exactly what happens. Bear's system basically predicts that Monday mornings between 9 and 11 AM are peak for mattress orders — people shop over the weekend and order on Monday. So the bots are already clustered near the mattress staging area by 8:45. They're not waiting for orders to drop. They're waiting for the pickers to show up. Luna: I love that operational foresight. It's a level of sophistication that ten years ago only Amazon could afford. Lucas: And now a mid-size DTC brand can get it for what amounts to a per-order fee. That's the democratization of logistics tech. Luna: Speaking of democratization — you know, it's interesting that conversations like this, where we get into the actual numbers and real case studies, are available to anyone who listens. A lot of this info used to be locked behind paid conference tickets. Lucas: Yeah, and that's part of why we do this show the way we do it — no ads, no paywall, just one focused topic per episode. A small group of listeners actually chip in monthly through buy me a coffee dot com slash fexingo, and that's what keeps these episodes coming. It's a pretty lean model — just listener support covering the costs. If these marketing conversations have sparked something you've actually used, that's the channel. Luna: It's a nice setup. Keeps us honest. No sponsor to please, just the topic and the listeners. Lucas: Exactly. So back to the robots — I think the next frontier for DTC brands is integrating the warehouse robot data with the ad platform data. Imagine if your fulfillment system tells your Facebook ads: 'We are currently processing orders in 2 hours, so increase bids for same-day eligible customers.' That's a feedback loop that doesn't exist yet but is totally feasible. Luna: That would be a game-changer for efficiency. You could dynamically adjust ad spend based on real-time operational capacity. If the robots are running at 80% capacity, you push more traffic. If they're maxed out, you pull back or shift to a later delivery promise. Lucas: Right. And some of the more advanced warehouse management systems already have APIs. Locus has an API that can push real-time throughput data. So a brand with a decent tech team could build that integration. I know of at least one pet food DTC brand that's experimenting with it — not ready to talk publicly yet, but they showed me a demo and it was impressive. Luna: So the next step isn't just using robots to fulfill orders. It's using them as a sensor network that informs every other part of the business — from ad spend to inventory planning to customer service response times. Lucas: Exactly. And that's where the marketing team needs to be in the room with the operations team. Too often those two departments don't talk. Marketing promises two-day shipping, but ops can't deliver it. Robots can close that gap. But only if marketing knows what operations can actually do. Luna: That's a good place to wrap. Any final takeaway for listeners who are considering warehouse robotics? Lucas: Start with your order volume and your current pick cost per order. If you're doing over 200 orders a day and paying more than $2 per order for picking labor, you can probably justify a pilot. Talk to Locus, 6 River Systems, or even GreyOrange. Ask for a trial with no long-term commitment. And track not just labor savings, but also the change in delivery speed and conversion rate. Those softer benefits might actually be the real ROI. Luna: Sounds like a plan. Thanks, Lucas. Lucas: Thanks, Luna. See you next episode.