We don't animate things to make them feel alive. We animate them so they reveal how they were built.
There's a kind of motion design that exists to entertain — page transitions that swirl, hover states that bounce, sections that part like the Red Sea. It's fun. It's also load-bearing on nothing. Strip it out and the page reads exactly the same. Whatever the user understood, they understood without it.
Motion as construction is different. It's an animation that shows you the relationship between two states: where the element came from, what it is now, what it can become. When a card expands into a detail view, you don't just see the new view — you see that the new view is made of the card. The hierarchy becomes visible.
This is hard to do well. Most motion frameworks make it easy to add a fade or a slide and call it done. The harder work is asking: what is this transition trying to teach? If the answer is nothing, cut it.
Test for it: mute every animation in your build. Does the meaning of the page survive? If yes, your motion was decoration. If no, you have construction. Most teams will find their answer somewhere in the middle, and the middle is the place to do the work.
Three rules we follow: motion should respect mass (heavy things move slower); transitions should reveal hierarchy (parents before children); a single signature moment per surface is more memorable than five small ones. Reduced-motion is a peer, not a fallback — design for it first.