Google is shutting down Doppl, its standalone virtual try-on app, on April 30. But don't read that as a retreat — it's the opposite. The company is embedding its AI-powered fitting room technology directly into Google Search and Shopping product listings, making virtual try-on a native part of the world's largest product discovery engine.

Starting April 30, anyone searching for clothing on Google can tap "Try It On" directly from product listings and image search results. No app download, no separate experience. Just search, tap, and see how a garment looks on your body — or on one of Google's diverse AI-generated model representations — before you buy.

The Technology Behind the Mirror

Google's system uses a custom diffusion-based image generation model trained on pairs of photos showing the same person wearing the same garment in different poses, according to reporting from HappyCapy Guide. The AI has learned how fabric drapes, folds, stretches, and casts shadows across different body types — the kind of physics-aware rendering that was science fiction five years ago.

The beauty integration is arguably even further along. Users searching for products like "pink lipstick" or specific SKUs can tap "try-on" to see shades live on their own face via their camera or on one of 148 diverse model faces, covering more than 50 beauty brands. Google has effectively become infrastructure for beauty AR, and most consumers don't even realize it yet.

Why This Matters: The $850 Billion Returns Problem

The timing isn't accidental. The National Retail Federation estimated that 15.8% of annual retail sales were returned in 2025, totaling a staggering $849.9 billion. For apparel specifically, return rates can exceed 30%. Every one of those returns represents a failed fit prediction — and a margin-crushing reverse logistics cost.

Virtual try-on technology has been the industry's white whale for over a decade. But as CNBC reported earlier this month, a growing wave of AI startups are finally making the technology good enough to meaningfully move the needle. The difference now is generative AI: instead of clumsy 3D avatars, modern systems produce photorealistic renderings that consumers actually trust.

Google embedding this directly into Search changes the calculus entirely. Retailers don't need to build their own try-on tools or integrate third-party solutions. If their products show up in Google Shopping — and virtually all of them do — they get virtual try-on for free.

The Competitive Landscape

Google isn't alone in this race. Amazon has been rolling out its own virtual try-on features, and startups like those profiled by CNBC are building specialized solutions for brands that want more control. But Google's distribution advantage is formidable: when try-on is one tap away from a search result, it becomes the default behavior rather than an opt-in feature.

The move also dovetails with Google's broader push into agentic commerce through its Universal Commerce Protocol. The vision is becoming clearer: Google wants to own the entire shopping journey from discovery through virtual fitting through checkout, all without the consumer ever leaving Google's ecosystem.

For retailers, the upside is obvious — fewer returns, higher conversion, less friction. The downside is equally clear: another layer of the customer relationship gets intermediated by Google. But with $850 billion in returns on the line, most will take the trade.