Motion isn't decoration: it's a sales argument

When a client asks us for «a bit of animation to make it feel more alive», we're almost always at the worst possible starting point. Animation isn't a coat of varnish applied at the end over a finished interface. It's a design decision with the same consequences as typography or the grid: it communicates something, whether or not there's intent behind it. The only real question is whether it communicates what we want.

We say it plainly because it's the hardest part to sell: motion is a commercial argument, not an ornament. Well-timed movement tells the user «someone who knows what they're doing made this» before they read a single headline. Gratuitous, sluggish or badly synced movement says exactly the opposite, and says it just as loudly. There's no neutral middle ground: the moment something moves on screen, it's already making claims about your product.

Perceived quality reads before the copy

There's a moment, between someone landing on a screen and starting to understand it, when they haven't read anything yet but have already formed a judgment. In that gap of milliseconds, motion does almost all the work. A transition that responds instantly, with a natural deceleration, reads as a solid, well-built product. The same screen with an animation that starts late, stretches too far or cuts off abruptly reads as fragile, even if the code underneath is identical. The quality a user thinks they see doesn't depend on how robust the system is, but on how it feels to touch. Motion is the main tool we have to govern that feeling.

Guiding attention instead of competing with it

Movement is the one visual stimulus the human eye can't ignore; our peripheral vision is wired to catch it. That makes it the most powerful and most dangerous tool in the kit. Used well, it steers the gaze to where it matters: an element that appears when the user completes a step, a state change confirming their action landed, a hierarchy revealed in the order it's meant to be read. Used badly, it hijacks attention toward the irrelevant: a carousel spinning on its own, an icon pulsing for no reason, six things moving at once forcing the eye to choose without knowing why. The question we ask of every animation isn't «does it look nice?» but «where is it sending the gaze, and is that where we need it?». If it doesn't lift conversion, reduce doubt or speed up a task, the movement is surplus: it's spending the user's attention and giving nothing back.

Timing and easing: the craft you don't see

Almost everything that separates a studio animation from a template one lives in two parameters nobody names: duration and curve. Duration has an honest, narrow range. Below roughly 150 milliseconds the change reads as an abrupt jump; above 400 the user waits and the interface feels slow. Most UI microinteractions live between 200 and 300 milliseconds, and a page transition should rarely exceed 400. Easing matters even more: real-world movement is never linear, it starts and stops with inertia. A curve that decelerates at the end (something close to ease-out) feels natural because it mimics how real objects come to rest. A constant speed, the default you get if you touch nothing, feels mechanical precisely because it doesn't exist in nature. Choosing these two numbers well isn't a finishing detail: it's the difference between something looking designed and looking like it just happened.

Nobody's going to tell you «great easing». They'll tell you your product feels expensive. It's the same thing said by someone who doesn't know they're saying it.

— Sors

Performance: a stuttering animation lies

All this intent collapses if the animation doesn't run smoothly. A transition that skips frames, that produces the stutter English calls jank, doesn't communicate quality: it communicates that something is wrong underneath, even if the user can't put it into words. And that judgment transfers to the entire product. Here design and engineering stop being two worlds: to hold 60 frames per second you have to animate only the properties the browser can resolve on the GPU without recalculating the page, essentially transform and opacity. Animating things like width, height or position via top and left forces the browser to redo where every element goes on each frame (what's called layout or reflow), and that's where the stutter appears. Translated into business terms: a gorgeous animation on paper that janks on a mid-range phone is worse than none at all, because the user you most need to convert is exactly the one who suffers it. Motion that sells is motion that performs on the client's real device, not on the designer's laptop.

Accessibility: prefers-reduced-motion isn't an extra

There's a group of users for whom movement isn't pleasant but literally uncomfortable or disabling: people with vestibular disorders for whom a parallax scroll or an abrupt zoom triggers real dizziness or nausea. Operating systems let people declare that preference, and the browser exposes it in a media query, prefers-reduced-motion, that we can check and honour. For us it isn't a compliance checkbox: it's part of doing the job well. Honouring it doesn't mean switching off all motion, it means having a second version in mind — swapping a large movement for a soft fade, dropping the bounce but keeping the state confirmation. The animation that confirms an action worked is still useful information even when reduced; the decorative part simply disappears. And there's a direct business argument, not just an ethical one: a flow that makes some of your users dizzy is a flow that loses those users. Accessibility and conversion, here, point to the same place.

When we defend a motion decision to a client, we don't talk about it looking pretty. We talk about perceived quality, directed attention, friction going down, and a slice of their audience we don't want to lose. That's the right frame, and it's what turns motion from an aesthetic indulgence into a budget line that justifies itself. Default movement, the kind you get from dropping in a component and deciding nothing, is always asserting something about your product. The only real choice you have is whether you let it speak for itself or hand it a script.