Why the feature list matters
Every serious platform: Threekit, Expivi, KBMax, Configure One, ShapeDiver, DriveWorks advertises roughly the same feature list. What separates a configurator that converts from one that stalls is not which boxes get ticked, but how well each feature is implemented on top of the 3D asset pipeline. A rules engine is useless if the mesh takes 12 seconds to load. AR is useless if the material looks plastic. CPQ is useless if the shopper closed the tab before the model finished streaming.
The 12 features below are the ones we treat as non-negotiable when we scope a build, and the ones we optimize the 3D asset pipeline around at Seven Studio.
The 12 essential features
- 1
Visual rules engine
A no-code editor where product managers define option groups, dependencies, exclusions, and price rules. Without one, every SKU change becomes a developer ticket.
- 2
Real-time 3D rendering
WebGL or WebGPU rendering that updates the model within 16 ms of every click. This is what separates a configurator from a static image gallery.
- 3
PBR materials and lighting
Physically based rendering with metalness, roughness, normal, and AO maps under an HDR environment. Anything less looks like a video game from 2010.
- 4
AR preview on mobile
One-tap AR through USDZ (iOS Quick Look) and GLB (Android Scene Viewer). This alone can lift conversion 40 to 90 percent on furniture and appliances.
- 5
CPQ pricing and quoting
Configure, Price, Quote logic that turns the current configuration into a live price, a downloadable PDF quote, or a Salesforce opportunity. No spreadsheet handoff.
- 6
ERP, PIM, and e-commerce integration
SKU generation, stock checks, Shopify or BigCommerce add-to-cart, and SAP or NetSuite order push. Without integration a configurator is a very expensive demo.
- 7
Shareable configuration URLs
Deep links (?config=abc123) that reopen the exact scene on any device. Sales teams and shoppers both need this: it is the single highest-ROI feature on the list.
- 8
Analytics and heatmaps
Event tracking on every option toggle, camera move, and drop-off. This is how you learn which materials sell and which options confuse buyers.
- 9
Multi-language and multi-currency
Localized option labels, units (mm vs inches), currency, and tax. A must for any brand selling into more than one region.
- 10
Admin and asset management
A back office where non-developers can add materials, swap textures, retire options, and preview changes before publish. Reduces engineering load by 80 percent over the product lifetime.
- 11
Performance and asset streaming
Draco or Meshopt compression, KTX2 textures, LOD swapping, and lazy loading of secondary materials. Target sub-3 MB initial payload and 60 fps on mid-range Android.
- 12
Accessibility and fallback
Keyboard navigation, screen reader labels, and a graceful 2D fallback for devices that cannot run WebGL. Ships you into enterprise procurement checklists.
Why every feature depends on 3D asset quality
Configurator software vendors sell features. Shoppers experience assets. A 200 MB CAD export dropped into a best-in-class platform will still stutter, look flat under HDR lighting, and fail on mobile. The same platform, fed a proper real-time asset: a retopologized hero mesh, PBR materials, KTX2-compressed textures, glTF with Draco or Meshopt: will render photorealistically at 60 fps on a mid-range phone.
- Real-time rendering only looks premium when the model has clean topology and a proper UV atlas.
- AR preview only convinces buyers when the material response matches how the product looks in daylight.
- Shareable URLs only work when the scene reloads in under three seconds, which is impossible with an unoptimized mesh.
- Analytics only tell the truth when performance is not the confounding variable: a slow model always drops off, regardless of the option set.
- CPQ and ERP integration land on the shopper's screen as a total price and an add-to-cart button. But the shopper only reaches that button if the 3D scene held their attention.
This is why we treat asset production as a first-class engineering discipline, not a handoff from marketing. Every feature in the list above is bottlenecked by the quality and weight of the mesh it renders.
A short buyer's guide
Use this rough matrix when evaluating configurator platforms against a custom build.
SaaS platforms (Threekit, Expivi, KBMax, ShapeDiver)
You want a managed back-end, a visual rules editor, and standard e-commerce hooks. Faster launch, ongoing per-seat or per-view license. Best under three product families and an eight-week deadline.
Custom builds (three.js, Babylon.js, model-viewer)
You need bespoke UX, unusual rules, hard performance targets, or full data ownership. Higher upfront engineering, zero license cost, unlimited creative control. Best with an in-house front-end team.
Frequently asked questions
What is product configurator software?
A visual commerce tool that lets a customer or sales rep build a valid, priced, manufacturable product by combining options. A modern one pairs a rules engine with a real-time 3D preview and pushes the final configuration to ERP or e-commerce.
What are the must-have features?
Rules engine, real-time 3D rendering, PBR materials, AR preview, CPQ pricing, ERP or e-commerce integration, shareable URLs, analytics, multi-language, admin tooling, and asset streaming for mobile performance.
Which platforms lead the market?
Threekit, Expivi, KBMax, Configure One, ShapeDiver, DriveWorks, Zakeke, and Cylindo cover most of the enterprise landscape. Custom three.js and Babylon.js builds cover the premium end.
How much does configurator software cost?
SaaS platforms range from a few hundred to several thousand dollars per month plus setup. Custom builds land between fifteen and eighty thousand dollars in engineering, with no ongoing license.
Why do 3D assets matter more than the software?
Every feature, rendering, AR, mobile performance, photoreal look, is bottlenecked by the mesh and materials the software renders. Underinvested assets make even a top platform feel slow and cheap.