On this page
Google has folded a try-on experience directly into Search, letting shoppers upload a photo and see how a garment drapes on their own body before they ever land on a brand's site. That's a genuine leap for product discovery. But a rendered image of a shirt on your torso doesn't tell you whether the size you're about to add to cart is the right one.
Search is where discovery starts, but it was never built to carry the last, most expensive decision in the funnel: which size to buy. That gap — between "this looks good on me" and "this will fit me" — is exactly where returns, bracketing, and margin erosion live.
What Google's Try It On Actually Does
Google's try-on tool sits inside Search and Shopping results. A shopper uploads a full-length photo, and Google generates a visualization of how a specific garment would look on that body, using a generative model rather than a simple overlay. It works across a large catalog of apparel because Google indexes product images at scale, and it requires no integration from the brand — the feature lives on Google's surface, not the retailer's.
That scale is the point. Google can show a plausible visual for millions of SKUs it has never touched physically, because it's solving a rendering problem: given this photo and this product image, produce a believable composite. It's worth noting Google also wound down its standalone Doppl try-on app in 2026 in favor of folding the capability directly into Search — a signal that this is meant to be a discovery-layer feature, not a destination.
The Difference Between Visualization and Fit
A rendering answers "what would this look like." A fit engine answers "which size should I order." These are different problems with different inputs.
Visualization needs a photo and a product image. Fit prediction needs the shopper's actual body measurements, the garment's real pattern dimensions, and how that specific brand's sizing runs relative to a standard — because a size Medium from one label can differ from a size Medium at another by multiple centimeters in the chest or waist. Google's tool, by design, isn't solving that second problem. It's showing you a garment on a photo-derived body shape, not cross-referencing your measurements against a brand's actual grading.
This distinction matters more than it sounds. Research cited by industry analysts consistently points to fit uncertainty, not style or color, as the dominant driver of apparel returns — shoppers order two sizes to hedge, or guess wrong and send it back. A beautiful visualization can actually increase purchase intent without reducing that uncertainty at all, which means a brand that only relies on Google's rendering could see more traffic and more orders, but no improvement in the return rate that follows them.
Why This Raises the Stakes on Owning the Fit Layer
Returns are already a structural cost problem, not a rounding error. US ecommerce returns topped roughly $890 billion in 2024, with online return rates running near 19–20%, according to the National Retail Federation. McKinsey has separately pegged the global cost of fashion returns at over $40 billion a year. Apparel and footwear sit above most other categories on return rate, and shoes are consistently among the worst offenders.
If Search sends a shopper to a product page having already shown them a flattering visualization, that shopper arrives with higher confidence in the look and no more confidence in the size than before. Left unaddressed, that's a setup for exactly the kind of bracketing behavior NRF has flagged as especially common among Gen Z shoppers — buying a Small and a Medium with the intent to return one. Google's feature can make the top of funnel more persuasive; it does nothing to close that specific gap, and it isn't designed to.
That's precisely why the fit layer — the part that recommends an actual size based on the shopper's body and the brand's real garment measurements — has to live with the brand, not with the search engine. Google doesn't have the brand's grading data, doesn't know how a specific fabric behaves, and has no incentive to reduce that brand's return rate. The retailer does.
Where YourSizer Fits
YourSizer operates at exactly this second layer — after a shopper has decided they like a garment, before they commit to a size. The widget takes a shopper's measurements (self-reported or derived from quick inputs) and cross-references them against the brand's actual size chart and fit data to recommend a specific size, with a virtual try-on visualization layered on top so the shopper sees both the look and the size recommendation together.
The practical difference from a generic Search-level try-on is that YourSizer's recommendation is brand-specific: it accounts for the fact that this particular brand's Large runs closer to another brand's Medium. That's the layer Google's feature structurally can't provide, because it isn't scoped to any one retailer's grading rules. For a brand thinking about where to invest next, the useful framing is that Google can improve how many people arrive interested; a size recommendation layer is what determines how many of those people keep what they ordered. Teams weighing this can look at how to build an internal business case for fit technology or review the ROI of a size recommendation tool before scoping a rollout.
Search-Level Try-On vs Brand-Owned Fit Tools
| Google Try It On | Brand-owned size + try-on widget | |
|---|---|---|
| Where it lives | Google Search / Shopping | Brand's product pages |
| Input | Shopper photo + product image | Shopper measurements + brand size chart |
| Output | Visual composite | Specific size recommendation + visual |
| Brand-specific grading | No | Yes |
| Reduces bracketing/returns | Not directly | Directly targets the decision point |
| Data ownership | Brand |
Frequently Asked Questions
Does Google's Try It On replace the need for a brand's own virtual try-on?
No. Google's feature is a discovery-stage visualization that helps shoppers imagine a garment on their own body while browsing Search results, but it isn't tied to a specific brand's sizing data. A brand still needs a fit layer on its own product pages to translate that interest into a correctly sized order, because the search-level tool has no way to know how that brand's particular Medium is cut.
Will Google's feature hurt or help brand conversion rates?
It should generally help top-of-funnel engagement, since a shopper who can preview a garment on their own body is more likely to click through with genuine interest rather than idle curiosity. The open question is what happens after that click. Without a brand-side fit recommendation, some portion of that added interest will convert into orders that get returned or exchanged, so the net effect on profitable conversion depends heavily on what the brand does on its own site.
Should brands stop investing in their own try-on tools now that Google offers one?
That would be a mistake for most apparel and footwear sellers. Google's tool operates at the platform level and is not customized to any single brand's measurements, fabric behavior, or grading quirks, so it can't function as a substitute for a brand-specific size recommendation engine. Brands that rely solely on Search-level visualization are effectively outsourcing the most persuasive part of the funnel while keeping full ownership of the costly part — returns, exchanges, and customer service tied to wrong sizes.
How does this connect to broader trends in AI and fit technology?
It fits a pattern where large platforms increasingly own the discovery and visualization layer, while brands retain responsibility for the transactional and fit-accuracy layer, similar to how AI-driven try-on built on diffusion models has moved from novelty to a mainstream expectation across ecommerce. Understanding how virtual try-on differs from size recommendation is useful context, since the two are often conflated but solve different problems in the funnel.
Conclusion
Google folding try-on into Search is a meaningful shift in how shoppers discover and evaluate clothing, and it will likely make product images across the web feel dated by comparison within a couple of years. But visualization and fit accuracy are separate problems, and only one of them is being solved by a platform that doesn't have access to a brand's actual sizing data. The brands that come out ahead won't be the ones that treat Google's feature as a replacement for their own fit layer — they'll be the ones that let Search do what it's good at, discovery, while keeping the size decision, and the return-rate consequences that follow it, firmly in their own hands.