How to Build a Virtual Showroom: The Real Production Steps

Before you build anything, settle one question: are you commissioning a custom space or assembling one yourself in a tool. A DIY platform like Zolak or Tailoor lets you upload products into a template and publish a 3D viewer on a monthly subscription, fast and cheap. A commissioned build, the kind we make on Babylon.js, is a bespoke space designed around one brand, starting from around 20,000 pounds plus monthly traffic and hosting. The 6 steps below describe the proper, bespoke version, because that is the one people usually mean when they ask how a virtual showroom is actually made, and because the DIY route mostly skips them by giving you a fixed template.

Here is what the work really involves.

First, choose how it will be delivered

One decision sits above the six steps, because it shapes all of them: how will people open the finished showroom.

In the browser (WebGL). The default, and what the steps below assume. The showroom is a 3D scene on a web page, reached on any phone or laptop with no app and no headset. We build on Babylon.js for exactly this reach: a customer is one tap from the space.

In a VR headset. The same production, output for hardware you control, a sales suite or a trade stand. Add it when you own the room; it is the wrong primary target for anything public.

AR on individual products. Not a room but a single item dropped into the user's space through their phone. It is a companion to a browser showroom, built from the same 3D assets, and it often ships first because it is the smaller job.

Settle this before step one, because a browser build, a VR build and an AR feature share assets but not scope.

1. Design the space, not a copy of your shop

Step one is not 3D at all. It is deciding how a visitor should move and discover. Good virtual showrooms are designed environments built around the products and the brand, not digital photocopies of a physical store. We plan the route, the focal points and the moments that matter, the same way a film set is built for a story. Skip this and you get a 3D room nobody wants to spend time in.

2. Build the product assets

Every product in the showroom needs a 3D model. These come from your existing CAD or manufacturing files, from photogrammetry, or built from scratch in software like 3ds Max with V-Ray for accurate materials. Those models are then converted to glTF or its compressed GLB form, the web standard for 3D, and optimised hard: reduced geometry, baked lighting, level-of-detail versions, so a 50 MB manufacturing file loads in 2 to 3 seconds on a phone.

This stage is where quality is won or lost. A showroom is only as convincing as its worst-looking product, and a brand that sells on design cannot afford a chair that renders like a video game prop.

3. Construct the scene in the browser engine

With assets ready, we build the actual environment in Babylon.js, the open-source WebGL and WebGPU engine. This is where the space comes together: lighting, materials, spatial layout, camera behaviour, and the rules for how a visitor walks through it. Babylon.js runs in any modern browser on desktop or mobile with no plugin, which is the whole point. A customer should be one tap from trying your showroom, not an app download.

4. Add the interaction

A 3D backdrop is not a showroom; interaction is what makes it one: hotspots that reveal product detail, the ability to turn and inspect an item, configurators that swap finishes, information panels, and the navigation between zones. This is the layer that separates a real virtual showroom from a flat catalogue with a 3D wallpaper behind it.

Deeper commerce hooks, live pricing pulled from your store and add-to-cart from inside the 3D space, are the hardest part and, in our case, in active development. Today a bespoke showroom drives discovery and routes the shopper to your existing checkout. Any studio should tell you plainly which interactions are ready now and which are roadmap.

5. Launch, host and measure

A finished showroom ships as a web package deployed on your own domain. It needs hosting that can serve 3D assets quickly, which is why a bespoke build carries a monthly traffic and hosting fee rather than a one-off price. Once live, you watch how visitors actually use it and refine, the same as any other part of your site.

What a studio needs from you

The build runs faster when the inputs are ready, and most delays are missing source material, not modelling time. In practice we need three things. The product data: CAD or manufacturing files where they exist, or clear reference for anything modelled from scratch, plus the real materials and finishes so a fabric reads as that fabric. The brand: your identity, the tone the space should carry, and any physical showroom or stand it should echo. And the priorities: which products lead, which configurations matter, and what a visitor is meant to do, so the interaction is built around real decisions rather than everything at once.

So how long and how much

A single-room showroom with one hero range is a small job. A multi-room flagship with dozens of configurable products is a large one, so timelines and cost scale with scope rather than a fixed menu. A bespoke build starts from around 20,000 pounds; SaaS costs a fraction of that but buys you a template, not a brand.

I will be straight about where we are: this is an emerging service for us. We have built five demo storefronts and a virtual furniture showroom, and we are developing the commerce integration now. If you want the honest version of what is possible today versus next year, that is exactly the conversation to have before you commit, and it is laid out in how we approach projects on how we work. For the wider picture of what these spaces are and what they cost, see our online showroom development.

Where the cost actually goes

A bespoke build is priced by effort, and the effort clusters in three places, which is why two showrooms at the same headline can cost very differently.

The 3D assets. The single biggest variable. A product with clean CAD files and simple materials is quick; a hand-modelled hero piece with complex fabric, glass or metal, optimised to load fast on a phone, is not. Ten products modelled properly costs more than a bigger room with five, so catalogue depth drives the number more than floor area does.

The interaction. A showroom with hotspots and spin-to-inspect is one price; add live configurators that swap finishes, information layers and zone navigation and the development grows. Deep commerce hooks into your store are the top of that scale and the least mature part of the category.

Hosting and traffic. Ongoing, not one-off. Serving 3D assets quickly to real visitors carries a monthly fee, which is why a bespoke build is priced as a project plus a running cost rather than a single invoice. When you compare a 20,000-pound custom build against a low monthly subscription, you are comparing an owned asset against a rental, not two versions of the same thing.

The mistakes that sink a build

Three recur. Modelling the whole catalogue in true 3D when a handful of hero products carry the story and the rest can be imagery, which burns budget for little return. Treating the space as decoration instead of navigation, so it looks good in a screenshot and confuses a visitor in motion. And skipping optimisation, shipping heavy models that stall on a phone, which quietly kills the mobile traffic the browser build was chosen to reach. A good studio designs against all three from the first step.

See a finished one

The clearest brief is a working example. We built an interactive virtual store that runs in the browser; walk the live build here to see what the steps above produce.

Frequently asked questions

What do you need from us to start?

Product files or clear reference, your real materials and finishes, brand assets, and a short list of which products and configurations matter most. The more of that is ready, the faster the build.

How long does it take?

It scales with scope. A single room around one hero range is a small job; a multi-room flagship with dozens of configurable products is a large one. It is a design and development project, so scope it like a website build with a 3D layer, not like a photo shoot.

Will it work on mobile?

Yes, when the assets are optimised properly. glTF models with reduced geometry and baked lighting let a heavy source file load in a couple of seconds in a phone browser. Skip that step and mobile is where it breaks.

Do we own the finished showroom?

With a bespoke build, yes: it deploys on your own domain and is yours like a website. A SaaS template is rented and disappears when the subscription stops.