The question that kills an online furniture sale is rarely “do I like this chair.” It’s “will this actually fit, and is it worth it” — asked silently, at 11pm, with no one around to answer. The newest addition to the Marvin XR viewer answers it anyway.
The gap between “looks nice” and “I’ll buy it”
3D and AR solved the first half of online furniture shopping: a customer can now rotate a sofa, drop it into their own living room, and see it at true scale. What that alone doesn’t solve is the second half — the specific, practical question that stops a hand from reaching for “Add to cart.” Is the frame solid wood or a veneer? Will it fit through a narrow doorway? What’s the actual seat depth, not just the marketing copy? Is it machine-washable? A shopper with one of these questions and no fast answer does one of two things: opens a live chat and waits, or closes the tab. Both cost you the sale that the beautiful 3D model just nearly won.
An AI assistant that lives inside the AR experience itself
Marvin XR campaigns can now include a conversational AI assistant embedded directly in the viewer — not a separate chat widget bolted onto the page, but part of the same experience the shopper is already rotating and inspecting. It answers the practical questions a product page and a spec sheet usually fail to: dimensions, materials, care instructions, delivery specifics, and anything else the merchant chooses to configure answers for. The shopper never has to leave the AR view or dig through an FAQ page to get an answer that’s specific to the exact product in front of them.
It sits naturally alongside the redesigned viewer controls already in the campaign — the real-world dimension overlay from AI Studio, and the Lighting, Background and Camera panel that lets a shopper see the product under different conditions before they decide. The assistant answers the question in words; the viewer around it lets the shopper verify the answer with their own eyes, in the same breath.
Why this is a returns story as much as a conversion story
Furniture and home goods carry some of the highest return rates in ecommerce, and a large share of those returns trace back to exactly the questions above — a size that didn’t match expectations, a material that looked different in person, a detail the shopper assumed rather than confirmed. An assistant that answers those questions honestly, before checkout, doesn’t just help close the sale; it closes the sale with a customer who actually knows what’s arriving. Fewer surprises at delivery means fewer returns, and a lower return rate is worth more to a furniture brand’s margin than almost any other single metric in the funnel.
Set once, answers forever
A merchant configures the assistant’s answers once per campaign — the real dimensions, the real materials, the real care and delivery details for that specific product — and every visitor from then on gets a consistent, accurate answer at 2pm or 2am, without a support queue in between. It’s the same instinct behind the rest of this round of updates: take the parts of selling online that used to require a human standing by, and make them work correctly on their own, so the humans on the team can spend their time on the leads and decisions that actually need a person.
Try it on a real, live campaign
This isn’t a mockup — it’s already running on a published Marvin XR campaign today, alongside the updated Lighting, Background and Camera controls and the real-world dimension overlay. Open it, rotate the product, and ask it the question you’d actually ask before buying.
The takeaway for a B2B buyer evaluating AR platforms: the bar has moved past “can it show the product.” The platforms worth shortlisting now are the ones that can also answer the customer’s next question, without a human on the other end — and price what that’s worth back to your team in the same dashboard.