When a 3D website will hurt your business

We build 3D websites and we would like to build yours. That is the disclosure. It is also why this post exists: plenty of the briefs that ask for 3D describe sites that would be worse with a scene in them, and saying so early is cheaper for everyone than discovering it in analytics.

Sites whose job is to be read

A blog. A documentation hub. A comparison page. A pricing page. Most B2B lead-generation pages. These pages have one job: get a specific piece of information in front of someone who is already looking for it, as fast as possible, and let them act.

A 3D layer on a page like that is friction wearing a costume. Even a perfectly optimised scene is something the visitor has to look past to reach what they came for, and the scroll-driven motion that makes a brand site memorable makes a reference page infuriating. If your best pages are the ones people read to the bottom, leave them flat and fast. Spend the 3D budget on the one page where you are trying to be remembered, not understood.

Audiences on low-end phones or bad connections

A 3D site can be built to load in two seconds on a mid-range Android — we do it, and phones get a still frame rather than a live scene. But “can be built” is not “usually is”, and even done properly, a still frame is a picture of an experience your mobile visitors are not having. If eighty percent of your traffic is on phones over mobile data, you are buying a desktop experience for twenty percent of your audience. That can still be right — if the twenty percent are the buyers — but it should be a decision you make with the traffic report open.

Checkout, sign-up and anything with a form

The closer a visitor gets to committing, the less anything on the screen should move. Motion near a form pulls attention from the fields; a scene behind a checkout competes with the thing you most want them to finish. The pattern that works is a 3D front door — the homepage, the product story — that hands off to plain, fast, still pages the moment someone decides to act. Any studio that proposes a 3D checkout is selling you their portfolio piece.

When the product has no shape

3D earns its cost when the thing you sell has a shape: a product that gets configured, a space people want to walk through, a piece of equipment worth turning over. It earns much less when the product is a service, a subscription or an idea. An abstract animated blob behind a consultancy’s homepage is decoration — it can be beautiful, and it can be right for a brand whose whole promise is being unlike the others, but it does not tell the visitor anything they did not know. Be honest about which one you are.

When nobody can maintain it

A 3D site has more moving parts than a flat one — a scene, an asset pipeline, fallback frames, a library that updates. If the studio hands it over and nobody on your side or theirs is going to touch it again, it will quietly rot: a browser change breaks a shader, a new product has no model, the fallback frame still shows last year’s colours. Before you commission one, know who owns it in year two.

What to build instead

The alternative to a 3D site is not a boring one. A fast, sharply designed flat site with real photography, considered type and motion used only where it explains something will outperform a mediocre 3D site on every metric that pays your bills — and it will cost less. We build those too, and we would rather build you the right one than the impressive one.

If the brief survives all five tests above, then a 3D site is probably going to do real work for you, and this is how we approach it. If it does not, you have just saved yourself the most expensive kind of mistake — the one that looks great in the launch announcement.

Sources and further reading

• Web Vitals — web.dev
• prefers-reduced-motion — MDN