DressX is teaming with Depop on a series of AI-powered resale styling and virtual try-on (VTO) activations, according to Retail Technology Innovation Hub. That is a more interesting development than another try-on launch. Resale is the hardest place to do try-on well, which is exactly why it is the best place to find out whether the technology actually works.
Start with why resale is so hard. On a brand's own site, every item comes from one catalogue, so a try-on tool has a tidy set of pieces to pull from and coordinate. Resale has none of that. The stock is a pile of one-off items from thousands of different sellers, in different sizes, conditions and styles. The whole appeal for the buyer is building a look out of mismatched, hunted-down pieces. Showing one secondhand jacket on a body, with nothing to put it with, answers the easy half of a question the shopper is actually asking with the hard half.
The wider momentum is easy to see. Forbes has covered DressX inside a wider Google-led try-on stack. Victoria Beckham has linked with DressX for virtual try-on. And DressX has published its own report laying out the benefits of the technology, which The Business of Fashion has turned into a piece on driving luxury conversion with AI-powered try-on. The signal across all of it is that shoppers want help seeing themselves in clothes before they buy.
Here is the point we keep coming back to. A try-on can answer two questions. The first is "does this item look right on me on its own?" A shopper looks at a dress on a body and decides whether they like it. Useful. The harder question, the one that decides whether the sale sticks, is "will this actually work?" Because clothes are never worn alone. The dress needs shoes. The blazer needs something underneath. The trousers need a top that does not fight them.
A shopper who loves a jacket on screen and cannot picture a single thing to wear it with is a shopper who closes the tab. And a garment can fit perfectly and still come back, because it did not go with anything they already own. Seeing one item on its own helps with the first question. The money is won or lost on the second.
We won't pretend DressX's report answers more than it does. It is built on DressX's own platform data, and DressX describes its findings as directional: a positive correlation between using try-on and exploring, adding to basket and buying. That is a signal worth having. It is not the same thing as proof that fewer parcels come back.
Glossy has run a sharp piece asking whether fashion's AI boom is solving a real problem. It is the most honest question in the category, and the answer depends entirely on what you point the technology at. Point it at a single garment and you get a nice preview. Point it at the real decision, the whole outfit, and you start touching the things brands actually care about.
This is the case for complete styled outfit try-on, which is what we build at Atomic Goats. Instead of dropping one item onto a model, we pull complementary pieces from across a brand's catalogue and show a full coordinated look, on a range of body types, in a setting that feels like real life. The shopper does not have to imagine the rest of the outfit. They see it. And it does not stop at the first order: Atomic Goats also builds looks around what a customer has already bought and sends them as a personal lookbook, so the next outfit starts from something they own.
That changes the decision. A coordinated look answers "will this work" before the parcel ships, which is exactly where style-driven returns tend to come from. It also gives cross-selling context, because the second and third pieces are styled in front of the shopper, as part of the look. Add optional personal colour analysis and you address colour too, which can be a factor in whether a garment suits someone, and which a studio photo of the garment alone says little about.
Resale is where this matters most, and where the DressX x Depop work is worth watching closely. With no unified catalogue and mixed sellers, the only thing holding a purchase together is whether the buyer can picture the look. If outfit-level try-on can prove itself in that mess, it can prove itself anywhere.
Read the report for what it is: shoppers want help. The whole look is the unit that sells, and the whole look is what we build.
Want to see your catalogue styled as complete outfits? Book a demo and we'll show you.