DaVinci Vaporizers — Case Study

DaVinci sells a modular device where the parts have to fit, and shoppers had no way to see what they were building. We built a guided configurator backed by 355 individual 3D models, so every valid combination renders in the browser before it reaches the cart.

Results

355 3D Models — One for every valid build combination, each loaded only when opened
69 → 96 Best Practices — Lighthouse score on the EQ product page across the engagement
Stock-aware Guided Build — Out-of-stock handling and add-to-cart blocking built into every step
5 Product Lines Rebuilt — EQ, ARTIQ, IQ3, IQ Core and MIQRO-C templates

The Challenge

The DaVinci EQ is a system of parts. A base engine, a heater, a crucible, a mouthpiece, an adapter and a bubbler, in finishes named Onyx, Sapphire, Gunmetal and Amethyst, and plenty of those parts do not fit each other. Shoppers were expected to assemble a several-hundred-dollar device from a static product page and work out compatibility on their own. They could not see what they were building, and they only found out a combination was invalid when it failed.The heatmaps showed where attention was going. On the EQ product page, image thumbnails took the largest share of mobile taps by a wide margin. On the homepage the hamburger menu did, which means people were navigating around the page on their own instead of being led through it. Shoppers were not reading the specs. They were trying to see the product.This storefront also carries constraints most do not. It is a regulated category, so every visitor clears an age gate before seeing anything, and the platform itself is a live risk: in mid-2026 Shopify announced it would suspend stores selling vaporizer products.

The Solution

We built the EQ configurator as a sequential builder. Nothing is pre-selected, each step opens once its prerequisites are chosen, and Start Over is always one tap away. Every step knows stock levels: unavailable options carry badges, incomplete or out-of-stock builds cannot reach the cart, and required adapters are added to the build automatically so the shopper never has to discover them.Behind it sit 355 individual 3D models, one for each valid combination, rendered live in the browser. The viewer is careful about weight, because a 3D experience that tanks the page helps nobody. WebGL and model files do not load until a shopper opens a 3D path, the renderer respects save-data and slow-connection signals, and rendering pauses whenever the model is off screen.Around the configurator we rebuilt the product templates for the whole line (EQ, ARTIQ, IQ3, IQ Core and MIQRO-C). Setup tabs route between the Jacuzzi, Skyrise and custom builds. Compatible-kit logic now lives in one shared component; it used to be duplicated per hero variant, which was letting accessory data drift between templates. Trust and shipping content moved below the buy area, where it supports the decision without delaying it. A 21+ age gate handles focus trapping, scroll locking and a remembered verification window.The system runs without us. Configurator labels and tooltips live in Shopify metaobjects, so DaVinci's own team edits the guidance, adds products and manages build options without touching code. When Shopify's vaporizer policy landed, we flagged it to the client before their own notice had reached them and took full data and theme backups the same week.