Latest / The DTC Marketing Podcast with Fexingo: Direct-to-Consumer Brands, Shopify, and E-Commerce Growth / How DTC Brands Use Augmented Reality Try-On to Cut Returns
Transcript
- Lucas: So you're shopping for a pair of sunglasses online. You've narrowed it down to two frames. But you can't try them on, and you're not sure how they'll look on your face. So you do nothing. Or you buy both, planning to return one. That's the classic DTC returns problem. Luna: And returns are brutal for DTC margins. We've talked about using returns data to improve product design, but what if you just stopped the return before it happened? Lucas: Right. That's the promise of augmented reality try-on. And in the last eighteen months, the technology has gotten good enough that some brands are seeing real, measurable hits to their return rates. Today I want to look at one specific case. Luna: Hit us with it. Lucas: There's a digitally native eyewear brand — I'm not going to name them because they're privately held and they don't share these numbers publicly, but I've talked to their head of e-commerce. They sell prescription glasses and sunglasses, direct to consumer, mostly through Shopify. Their return rate was hovering around 12 percent, which is actually decent for eyewear. But every return costs them about $8 in shipping and restocking. Luna: And eyewear returns are especially tricky because the customer might just not like the fit on their face. It's not a defect; it's a preference. Lucas: Exactly. So in early 2025, they rolled out an AR try-on feature on their product pages. You click a button, it accesses your phone camera, and it overlays the frames onto your face in real time. The technology came from a company called Banuba, which is one of the bigger players in this space. The integration took about three weeks of developer time. Luna: Three weeks? That's faster than I would have guessed. Lucas: It helped that they were on Shopify, which has a decent AR kit plugin. But the key decision was that they didn't just slap it on the site and forget it. They actually redesigned the product page so the AR try-on button was above the fold, next to the main product image. And they added a short video tutorial showing how to use it. Luna: We deliberately don't run ads on these shows. If you want to support that choice, the link is buy me a coffee dot com slash fexingo. And honestly, if these marketing conversations have sparked something you've actually used, that's the best kind of support. Lucas: Yeah, exactly. It keeps the show ad-free and focused on what actually works. So back to that eyewear brand — within three months, their return rate dropped from 12 percent to 9.8 percent. That's an 18 percent reduction in returns. Luna: That's real money. On a million dollars in revenue, that's saving tens of thousands in shipping and restocking costs alone. Lucas: But here's the part I found surprising. The conversion rate on product pages where customers used the AR try-on was 27 percent higher than pages where they didn't. And the average order value was also about 8 percent higher. Luna: So AR didn't just reduce returns. It actually helped people buy more confidently, and they spent more. Lucas: That's the thesis. And it's playing out across categories. I've seen similar numbers from a DTC furniture brand that uses AR to let customers place a virtual sofa in their living room. Their return rate on ar interacted products was 22 percent lower than on non ar products. Luna: Furniture returns are even more expensive because of freight. So the savings there are huge. But I'm curious about the technical barriers. A lot of DTC brands are small teams. Can they really pull this off? Lucas: The cost has come down dramatically. Three years ago, you'd need a dedicated 3D modeling studio. Today, you can generate a usable 3D model from a series of smartphone photos. There are services like VNTANA and CGTrader that will create a model for under $200 per SKU. Luna: That's per product, though. If you have five hundred SKUs, that's a hundred thousand dollars. That's not nothing. Lucas: True. But the ROI math works if you have high-return categories. The eyewear brand I mentioned had about 80 frame styles. So their 3D modeling cost was roughly $16,000. They recouped that in about four months just from reduced return shipping costs. Luna: And then there's the ongoing benefit of higher conversion and AOV. So it becomes a no-brainer after a certain scale. Lucas: One thing I want to flag: the technology isn't perfect. It works best on newer phones with good depth sensors, which is about 70 percent of smartphone users in the US as of mid-2026. On older phones, the experience can be glitchy. So you have to decide whether to gate the feature behind a device check or let everyone try it and risk a bad experience. Luna: What did the eyewear brand do? Lucas: They opted for a progressive enhancement approach. If the phone supports AR, they show the button. If not, they show a static 3D rotation view that still gives a better sense of the product than a flat image. They saw about 40 percent of their mobile traffic actually using the AR feature. Luna: That's a decent adoption rate. And I imagine it's growing as people get more comfortable with the idea. Lucas: Absolutely. The cultural shift is real. Remember when face filters on Instagram felt weird? Now it's normal. AR try-on is just a commercial version of that muscle memory. Luna: So what's the catch? I keep waiting for the downside. Lucas: The biggest catch is that AR try-on can actually increase returns if the model isn't accurate enough. If the virtual product doesn't match the physical product in color, scale, or fit, customers feel deceived and return even more. There's a well-known cautionary tale of a fashion brand that used low-quality 3D models and saw their return rate spike by 6 percent. Luna: So accuracy is everything. You can't just put a rough overlay and call it a day. Lucas: You need to validate the models against the real product. The good AR providers will do a calibration step where they compare the virtual object to the physical one under different lighting conditions. The best brands also A/B test the AR experience before rolling it out sitewide. Luna: What about the impact on customer support? Does AR reduce the number of 'does this fit?' emails? Lucas: The eyewear brand told me that their pre-purchase customer service inquiries dropped by 14 percent. People were asking fewer questions about frame size and style. But post-purchase inquiries about fit actually went up slightly, because some customers thought the AR was a perfect simulation and were disappointed when the real product felt different on their face. Luna: So AR sets a higher expectation. You have to manage that with clear copy: 'This is an approximation, actual fit may vary.' Lucas: Exactly. And some brands are adding a disclaimer that shows a confidence score for the AR match. That actually builds trust, because it's honest about the limitations. Luna: I want to zoom out a bit. We're mid-2026. Is AR try-on becoming table stakes for DTC? Or are early adopters still getting an advantage? Lucas: I think we're in a window. The big platforms — Shopify, BigCommerce, Salesforce — are all building native AR capabilities. So within a year or two, it'll be a standard feature like product zoom. But right now, only about 15 percent of DTC brands have implemented it. The rest are watching. Luna: That's a classic early-mover opportunity. But it requires investment and a willingness to iterate. Lucas: And the brands that do it well are seeing a compound effect. Lower returns, higher conversion, better customer satisfaction. One mattress company told me that their AR try-on — which lets you visualize the mattress in your bedroom — increased the time spent on the product page by 2.5 minutes, which gave them more opportunity to upsell the pillow and sheets. Luna: That's interesting. AR is a discovery tool, not just a decision tool. Lucas: Right. It changes the browsing behavior. People start exploring products they wouldn't have considered because they can now see how they'd look or fit. That's why AOV goes up. Luna: Okay, so if I'm a DTC marketer listening right now, what's the one concrete step I should take this week? Lucas: Pick your three best-selling products. Get a high-quality 3D model made for each. Then run a two-week A/B test on the product page: one version with a static image, one with an AR button. Measure not just conversion and returns, but also time on page and bounce rate. That'll give you the specific data you need to justify scaling the project. Luna: And if the test works, you have a clear story to tell your CFO. Lucas: Exactly. The numbers are there. The technology is ready. It's really about execution and being honest with customers about what the tool can and can't do. Luna: I think that's a solid takeaway. We'll be watching to see which brands nail this over the next year. Lucas: And we'll report back. For The DTC Marketing Podcast, I'm Lucas. Luna: And I'm Luna. See you next time.