Newsroom

Sports equipment has a particular online problem. A lot of it is large, most of it is bought without being seen, and the two questions a shopper cannot answer from a photograph are where will it go and how big is it really. A rowing machine is 2.4 metres long. A road bike needs a hallway it can actually turn in. A kayak on a rack changes a garage. None of that is visible in a product shot on a white background.

This is the situation augmented reality is genuinely good at, and it is worth separating from the try-on conversation that dominates AR coverage. Placing a correctly scaled object in a room is a solved, reliable, boring piece of technology. Simulating how a running shoe fits a foot is not.

The space question

For anything over about a metre in any dimension, the shopper's real objection is not the product. It is the room. Will this fit between the sofa and the wall. Can I still open the door. Where does it live when nobody is using it.

A product page can answer that badly with a dimensions table, or well by letting the shopper stand the thing in their own space through their phone camera. The second is one tap from the page, needs no app, and takes about fifteen seconds. It replaces a tape measure and a guess with something the shopper can walk around.

The categories where this matters most:

• Cardio equipment. Treadmills, rowers, bikes, ellipticals. All large, all bought for a specific corner of a specific room.
• Racks, benches and home gym frames. Height clearance is the usual failure and the usual return.
• Bikes. Less about the room, more about storage and transport, and about whether the frame is the size the shopper pictured.
• Outdoor and camping. Tents especially, where the packed size and the pitched size are two very different objects and shoppers only see one of them in the photos.
• Table games and large recreation gear. Table tennis, pool, basketball hoops, trampolines.

The fit question, and where AR stops

Fit for wearable sports gear is a different problem and AR is much weaker at it. Helmets, shoes, gloves and packs all depend on the shape of a body part measured to the millimetre, which a phone camera does not do reliably. Face tracking is good enough for the silhouette of a helmet or a pair of sports sunglasses, and that is real value, but it is not a sizing tool.

What does work for wearables is a well made 3D model with proper zoom. A shopper looking at the sole pattern, the strap adjustment or the vent layout at close range is getting information a photo set rarely provides, on every device, with no camera permission. Build that first.

What the evidence actually supports

There is no good published dataset specific to sports equipment and AR, and anyone quoting one should be asked for the link. What exists is general and points the same direction. Shopify's merchant data from 2020 found that interactions with products carrying 3D or AR content converted 94 percent more often than the same kind of products without. Gunner Kennels, which sells a large hard-sided product where size in the room is the entire purchase question, reported a 5 percent reduction in returns after adding AR. That product profile is much closer to a rowing machine than a t-shirt is.

Returns matter here more than in most categories because the freight is expensive. The National Retail Federation put returns at 14.5 percent of United States retail sales in 2023, with online returns running at 17.6 percent. A returned treadmill is not a postage label, it is a pallet.

What it takes to build

The mechanics are the same as any other category and are not complicated. You need an accurate 3D model per product, exported as GLB for the web and Android and USDZ for iOS, hosted on a CDN, and a viewer component on the product template. Google recommends keeping Scene Viewer models under 10 megabytes with textures at 2048 by 2048 or below, which is comfortable for most equipment as long as the model is built for the web rather than exported straight from CAD.

Large equipment has one specific modelling trap worth naming. Products with a folded and an unfolded state, which is most cardio equipment and every tent, need both states. A treadmill that only exists in the running position does not answer the question the shopper actually has, which is how much space it takes when it is put away. Model both and let the shopper switch.

Where to start

Take the ten products with the highest return rate or the most pre-sale questions about size. Those are almost always the large ones. Build models for those, put a 3D viewer and an AR button on their pages, and compare their returns and conversion against the rest of the catalogue over a full quarter. That is a real test, and it costs a fraction of doing the whole range.

We can build the first one for nothing so you can see the quality and put it on a page before deciding. Send us a product with photos and measurements. Free sample here. If you want the full picture of formats, hosting and what it costs per product, that is on our solutions and prices pages.

Sources

• Shopify, merchant data on 3D and AR content, September 2020.
• Shopify, AR shopping, including the Gunner Kennels returns result.
• National Retail Federation and Appriss Retail, 2023 Consumer Returns in the Retail Industry.
• Google, Scene Viewer model guidelines.