When the client asks for animation

Most animation requests are requests for the wrong thing.

The pattern, when a client asks for “animation” on their site, is recognisable. They have seen another site, recently, with motion that looked considered. They want their site to look that considered. They translate the request into “we want animation”. The studio is then asked to add animation, which is the symptom, not the cause.

A few questions the studio now asks before any animation work starts.

What did the other site do that you liked? Specifically. Was it the page transition. The hover effect on a card. The way the headline crossfaded into a different headline on scroll. The way an illustration moved when it came into view. Most clients cannot answer this without going to the site and showing me the section. The act of showing me usually narrows the request to something specific.

What is the animation meant to communicate? If the answer is “it makes the site feel more polished”, the animation is doing a brand job, not a UX job. The brand job is, much of the time, achievable without animation. Better typography, better spacing, better photography. Animation is a slow, expensive way to add polish that other levers can add faster.

What is the animation hiding? Most loading animations are hiding slow loads. Most transition animations are hiding slow page changes. Most micro-interactions are hiding the absence of feedback. The fix, in each case, is to address the underlying problem (page weight, server response time, feedback signal) rather than to layer animation on top.

The animation work the studio does ship, after this conversation, tends to fall into a small set of categories.

State transitions that have a clear before and after. A modal opening. A menu sliding in. A toggle switching state. The animation here is signalling causation. The user did a thing. The thing happened. The animation makes the connection visible. Useful.

Loading states for genuinely slow operations. Skeleton screens, progressive image loading, indeterminate spinners. The animation here is occupying time. It is doing work the user cannot see. Useful, when honest.

Decorative motion that is part of the brand identity. The wordmark that animates on first load. The illustration that breathes when the page is idle. Useful, in small doses, on brands where the motion is actually doing brand work.

The animation work the studio declines, after this conversation:

Scroll-triggered fade-ins on every block. Almost never useful. The user is scrolling. They see the content. The fade-in delays the moment when the content is readable. The animation is a hindrance.

Hover effects that change layout. Almost never useful. The user is moving the mouse. They do not want the page to redraw under their pointer.

Page transitions that are slower than the page change would be without them. Almost never useful. The user is here to read the next page, not to watch the transition.

The single resource on motion that has shaped how I think about it most is Val Head’s writing, particularly her work on the timing curves that signal different kinds of state change. The book Designing Interface Animation is the deepest treatment of the topic and aged better than most of the design writing of its era.

Rachel Nabors on motion design for the web and the MDN entry on prefers-reduced-motion cover the accessibility side of the same conversation, which the design press skips more often than it should.

Most animation requests, if the conversation is run carefully, end with the client agreeing that what they wanted was not animation. They wanted a site that felt considered. The animation was the proxy. Ship the considered site. The animation, where it is needed, will be obvious.