3D configurators: when they help and how to measure it

Someone lands on an armchair product page. They like the design but cannot choose between fabrics or see the back. Another photograph might answer the second question; a configurator might answer the first.

Choosing a 3D feature starts with separating those needs.

Viewer, configurator or augmented reality?

A 3D viewer lets people explore a model. A configurator changes options such as materials or components and must respect combinations you can sell. Augmented reality, or AR, places a representation in the shopper's surroundings.

They can coexist, but evidence about an AR experience does not establish how much any configurator will sell.

Reading success stories carefully

Shopify reports that Rebecca Minkoff visitors who interacted in AR were 65% more likely to order. That group may have different purchase intent. The published figure alone does not establish an equivalent causal lift across all visitors.

Gunner Kennels addresses a recognisable problem, choosing a crate size, and compares conversion rates with the previous quarter after adding 3D and AR. It is useful for forming a hypothesis, rather than a guarantee for another catalogue.

Research by Tan, Chandukala and Reddy associates AR use with sales in a cosmetics app, particularly where product uncertainty is greater. That setting does not represent every use of 3D either.

Ask which buying uncertainty your experience can resolve before choosing a percentage to justify it.

Choose one product for a pilot

Look for a product with recurring questions about shape, finishes or dimensions and enough traffic to learn from. “The finish wasn't what I expected” suggests a different problem from “It arrived late”.

Check that variants can be represented accurately and decide who will maintain the content. New finishes may require materials, reference images, visual approval and catalogue integration.

The pilot should test that operational work as well as the shopping experience.

Decide what to measure

Write a hypothesis before building: “Showing combinations will help shoppers choose a variant and complete their order.” Choose a primary measure and checks for unwanted effects.

Where feasible, randomly assign visitors to the usual page or the configurator version. Analyse everyone assigned, rather than only people who interacted.

Agree on duration, sample size and the smallest meaningful change with whoever manages analytics. Allow time for returns. With low traffic, usability sessions and real support questions can provide learning without turning a handful of sales into a firm conclusion.

Keep the product page useful without 3D

Provide photographs, dimensions and accessible options from the start. Check that the selected variant reaches the cart correctly and that buying remains possible when the model loads slowly or a shopper prefers not to open it.

Test with a keyboard and representative phones. Our guides to store accessibility and mobile 3D performance develop those checks.

To plan a pilot, tell us about your product and the questions buyers ask.