There are roughly 15 million digital screens in retail environments worldwide. Most of them are playing slideshows.
Nicely designed slideshows, sure. Professional-grade content management systems. Perfectly timed loops. But at the end of the day: static content on an expensive screen that nobody can measure, nobody can interact with, and nobody can prove drives a single sale.
We think that is an extraordinary waste of perfectly good hardware.
This article is about what happens when you turn those screens into something people actually want to stop and use. And why the same API that powers virtual try-on inside a mobile app can power an AR mirror in a flagship store, a kiosk in a shopping mall, and an interactive storefront window — without a single line of platform-specific code.
If you work in enterprise retail, digital signage, or you are the developer who will eventually have to integrate this: this one is for you.
The digital signage industry is worth tens of billions. Companies have spent the last two decades putting screens everywhere: malls, airports, transit stations, store windows, fitting rooms. The hardware keeps getting better. The content management keeps getting slicker.
And the value proposition has stayed more or less the same since 2005: put an image on a screen. Maybe a video. Loop it. Hope someone looks.
That is not a criticism of the signage companies — they have built impressive infrastructure. Reliable hardware, global deployment capability, serious network management. The problem is that this infrastructure is dramatically underutilised. It is the digital equivalent of buying a Ferrari and using it exclusively for the school run.
Every retailer we talk to has screens. Most of them cannot tell you whether those screens have ever influenced a purchase.
In digital advertising, everything is measured. Click-through rates, conversion rates, ROAS, attribution models that would give you a headache. In digital signage? The best most operators can offer is: impressions.
Impressions means: a screen was on, and people probably walked past it.
Try selling that to a brand that spends six figures on a seasonal campaign. They want to know what happened. Did anyone engage? Did anyone buy? Can you prove it?
This gap between what digital signage could measure and what it actually measures is the reason the industry is stuck selling hardware with declining margins. The screens are everywhere. The data is nowhere.
Here is where it gets interesting.
A screen with a camera and a virtual try-on API becomes something fundamentally different from a screen playing a slideshow. It becomes a destination. A thing people walk up to and use. A thing you can measure.
We keep calling them magic mirrors because the name stuck, but let us be honest about what they are: a screen, a camera, and a web browser.
That is it. No special hardware. No proprietary operating system. No custom drivers. A standard display with a standard camera running a standard browser.
Our virtual try-on engine runs inside that browser. A customer walks up, the camera picks them up, and they can try on shoes, bags, or accessories in real time. Same rendering quality as the mobile app. Same tracking precision. Because it is the same API.
The mirror does not know it is a mirror. As far as our system is concerned, it is just another browser window requesting try-on sessions. Which is exactly the point.
For signage companies, this is important: you do not need to build anything special. If your hardware runs a browser — and in 2026, it does — you already have the infrastructure to deploy interactive virtual try-on.
This is the use case that surprises people the most.
Picture a fashion brand’s storefront at 10pm. The store is closed. But the window display has a screen, a camera behind the glass, and our API running.
A person walking by sees the latest collection. They stop. The screen detects them through the glass and shows a virtual try-on experience. They see themselves wearing the shoes. A QR code appears. They scan it with their phone, and the session — the exact products they just tried on — transfers to their mobile browser. They can add to cart, save to wishlist, or buy right there on the pavement.
The store is closed. The screen is selling.
For signage operators, this is a completely different conversation with retailers. You are not selling a screen that plays ads. You are selling a 24/7 sales channel that captures leads from foot traffic and can prove it with data.
Mall operators have a particular problem: they have common-area screens that tenants pay for, but the value proposition is vague. Brand awareness. Foot traffic direction. Nice logos.
Now imagine a kiosk in the main atrium where shoppers can try on shoes from five different tenant brands. Each try-on session is logged. Each QR handoff is tracked. Each downstream purchase is attributable.
That kiosk just became a retail media product. The mall can sell interactive try-on slots to tenants at a premium over static ad space, because the data proves engagement and conversion. Not impressions. Engagement.
If you are the developer reading this and thinking “okay but what does integration actually look like,” this section is for you.
Our system is a hybrid architecture we covered in detail last week. The short version:
Server-side rendering. The heavy computation — body tracking, 3D rendering, product overlay — happens in our pipeline. Not on the device.
Web-based client. The client handles the camera feed and displays the result. It runs in any modern browser. No native code.
Hardware-agnostic. If it has a camera and a browser, we can run on it. A phone, a tablet, a 55-inch retail display, a mirror with a computer behind it, a kiosk at an airport.
The integration surface is the same regardless of the endpoint. You are calling the same API, using the same WebSocket connection for the real-time feed, and rendering in the same way whether your target is a mobile app WebView or a full-screen Chrome instance on a digital signage player.
One set of docs. One support channel. No per-platform surprises.
If you are a signage company evaluating this: your developers integrate once. They do not need to learn a new framework. They do not need to manage a native SDK. They connect to our API, handle the camera input, and render the output.
The same integration works across every screen in your network. Different screen sizes, different orientations, different locations — same code. You are not maintaining N integrations for N hardware configurations. You are maintaining one.
For retailers deploying across channels, it means the VTO experience in the mobile app, on the website, and on the in-store mirror is literally identical. Not “similar.” Identical. Because it is the same rendering pipeline.
Developers always ask about connectivity. Fair question. A mall has great WiFi. A bus shelter does not. A pop-up store at a festival might have nothing.
The core rendering engine and product catalog can operate locally on the device. That means the try-on experience works even with unstable or no internet connectivity. The session data syncs when a connection is available.
For signage deployments in transit stations, basement retail, outdoor pop-ups, or anywhere connectivity is unreliable: this matters. Your screen does not go dark because the WiFi dropped.
This is the part that ties the whole omnichannel story together.
A customer interacts with a try-on screen in-store. They find three pairs of shoes they like. A QR code appears on the screen. They scan it with their phone. Their session — the exact products, the exact views — transfers to their mobile browser.
Now they are on their phone, inside the retailer’s mobile experience, with the products they discovered on the screen already loaded. They can continue trying on, save to wishlist, or purchase.
The screen started the journey. The phone continues it. The data connects them.
For retail media measurement, this is the missing piece. You can now track the full funnel:
Impressions — how many people walked past the screen (the signage operator’s world).
Interactions and virtual try-ons — how many people stopped and engaged (our world).
Downstream purchases and ROAS — how many people bought after scanning the QR code (the retailer’s world).
Full-funnel attribution from a screen in a mall to a purchase on a couch. That is what turns a digital signage screen into a retail media product.
If you are a digital signage company, you know the margin pressure. Hardware is commoditising. The screens themselves are cheaper every year. Competing on hardware specs is a race to the bottom.
Interactive VTO changes the economic model. The screen is no longer just hardware you sell once. It is a platform you sell with recurring software and media revenue.
A “dumb” screen sold for its panel quality competes on price. A smart screen sold with built-in virtual try-on, session analytics, and retail media capability competes on value. That is a fundamentally different sales conversation.
You can justify higher hardware margins because the screen comes bundled with software that generates measurable ROI. And the software generates recurring revenue — either from the retailer directly or from brands buying CPE-based media slots on the screen.
CPM (cost per thousand impressions) is how most digital signage media is sold today. The problem: impressions are barely a step above guesswork in physical environments.
CPE (cost per engagement) is what interactive screens make possible. A try-on session is an engagement. A QR scan is an engagement. A downstream purchase is an attribution.
Brands will pay significantly more for CPE than CPM because the data actually proves something happened. A fashion brand running a shoe launch on an interactive screen can see exactly how many people tried the shoes virtually, how many scanned the QR code, and how many purchased. That is a media product worth paying premium rates for.
One of the world’s largest fashion groups deployed our VTO across their stores using this exact model. AR mirrors in flagship locations, running the same API that powers their mobile app and website.
A customer walks into a store in Madrid, tries on shoes using an AR mirror, scans a QR code, and later that evening finishes the purchase from their phone at home. Same session. Same products. Full attribution.
The signage hardware was standard. The integration was web-based. The VTO experience was identical to mobile. And the data connected everything.
They did not need three separate systems. They did not need three separate teams. One pipeline. Mobile, web, in-store — same code.
Whether you are a signage company looking to add interactive capability to your product line, a retailer evaluating omnichannel VTO, or a developer scoping the integration:
1. Pick your highest-impact use case. A footwear mirror in a flagship store. An interactive window on a high-street location. A kiosk in a mall atrium. Start where the foot traffic is and the measurement gap is largest.
2. Check your existing hardware. If your screens have cameras and run a browser, you probably do not need new hardware. The API is hardware-agnostic.
3. Think about the data story. The VTO is not the end goal. The data is. Engagement metrics, session handoff, full-funnel attribution — that is what makes this a media product instead of a feature demo.
4. Start with one screen. Integrate, deploy, measure. If the engagement and conversion data tells the right story, scaling across your network is the same code deployed to more endpoints.
We have done this before, and we are happy to walk you through it. Book a demo and bring your platform team — or your signage partner. The best conversations happen when hardware and software are in the same room.
• • •
The digital signage industry built incredible infrastructure. Screens everywhere. Global networks. Reliable hardware. Serious operational capability.
What it did not build is a reason for anyone to stop and interact.
Virtual try-on changes that. One API turns a passive display into an interactive experience that people actually want to use. The same API that runs on mobile and web, so there is no separate integration to maintain. And the session handoff from screen to phone means the data finally connects the in-store experience to the online purchase.
Screens do not have to be slideshows. They can be try-on points, revenue centres, and measurable media products.
The hardware is already there. The infrastructure is already there. The missing piece was the software layer that makes people stop, engage, and buy.
That is what we built.