Craft·June 2026

Motion Graphics That Don't Look Like Templates

There's a text animation I can spot from across a room. The title slides up, overshoots a little, settles back with a soft elastic bounce. You've seen it. Everyone has seen it. It shipped in a preset pack that half the internet downloaded a few years ago, and it now lives in fitness reels, finance reels, cooking reels, podcast clips, real estate walkthroughs, and at least one wedding invite video someone forwarded me last month.

I know that pack well.

We owned it too.

That's the uncomfortable place this essay has to start, so let me just say it plainly: in the earliest days of With Media, we used template packs. Bought some, downloaded the free ones, dragged the presets onto client footage and called it motion graphics. I'm not going to pretend we arrived fully formed with strong opinions about typography. We arrived broke and in a hurry, the way everyone does. I was undercharging, saying yes to everything, and a preset that made text bounce in two clicks felt like a gift.

What changed my mind wasn't a design course. It was a scroll session.

The night I couldn't tell our own work apart

I have a bad habit. When a competitor's reel bugs me, I'll watch it on loop late at night, trying to name what exactly is bothering me. It's not healthy, but it's honest research. One of those nights, deep in a feed of coaching content, a reel came up that I was fairly confident we had edited. Same title animation we used. Same transition into the b-roll. Same little subscribe nudge in the corner.

I checked. It wasn't ours. It wasn't even a client in our category.

That's the whole problem compressed into one moment. If I — the person who runs the studio, who has sat through the edit of well over a thousand videos — cannot tell our work from a stranger's at a glance, what chance does a viewer have of recognising the client? None. And recognition is the entire game. On a feed, you have maybe half a second before someone decides whether this is a video from a person they know or just more content.

Templates don't make you look bad. That's the part people get wrong about them, and it's why the argument against them usually fails. Most commercial packs are competently made. The animation curves are smooth, the timing is fine, the renders are clean. Nothing about them is bad.

They make you look like everyone else. Which, in an attention economy that runs on recognition, is quietly worse than looking bad. A slightly rough video that is unmistakably yours beats a polished video that could be anyone's. I believed the opposite for at least the first year of running this studio, and it took losing that late-night game of "is this ours?" to fix it.

Templates don't make you look bad. They make you look like everyone else — and in a feed, that's worse.

Motion has a job, and it isn't decoration

Here's the reframe that actually changed how we work, and it has nothing to do with taste: motion graphics have a job. Decoration is not the job.

the four jobs a graphic must earn its place doing.

The job is directing attention.

Every frame of a video is a small argument with the viewer's eye. Where should you look? What matters right now? What just changed? Good motion design answers those questions before the viewer even knows they were asked. When I break it down for new editors joining the team, I tell them motion earns its place in the timeline by doing one of four things.

It guides the eye. A frame has a hierarchy whether you design one or not. If your client is talking and a graphic appears, the viewer's eye has to choose. Motion decides that choice for them — it pulls focus to the thing that matters at the moment it matters, then hands focus back. An arrow that draws itself toward the number on screen isn't decoration; it's a hand on the viewer's shoulder saying look here now.

It emphasises the word that matters. Speech has stress. When someone says "forty grams of pure veg protein in twenty minutes," the sentence has a shape — certain words carry the claim, the rest carry the grammar. On-screen type should mirror that stress. When we cut "40g Pure Veg Protein in 20 Mins," the whole video is the number. If the type treatment gives "40g" the same visual weight as "in," you've flattened the one thing the video exists to say.

It visualises what footage can't show. A camera can film a person, a kitchen, a workout. It cannot film a comparison, a timeline, a before-and-after gap, or the idea of seventy-three days of slow progress. When we made "Average Indian Learns the World's Hardest Pushup in 73 Days," the footage could show the attempts — the failure, the shaking arms, the eventual rep. But the sense of time, the weight of that number, the feeling that day 40 is very far from day 1? That lives in graphics. Numbers, dates, progression — rendered so you feel the distance rather than just hear it.

And it bridges cuts. Every edit is a small jolt. Sometimes you want the jolt — a hard cut is a percussion hit. But often you want the energy of one shot to carry into the next, and motion is how you do it: an element that travels across the cut, a wipe that follows a gesture, type that persists while the world changes behind it. Done well, the video feels directed. Done badly — or not at all — it feels assembled.

That's the whole test we run now, on every graphic in every timeline: which of the four jobs is this doing? If the answer is "none, but it looks nice," it comes out. Not because minimalism is fashionable. Because a graphic with no job is stealing attention from a graphic with one.

The first betrayal: everyone bought the same pack

So why do templates fail at this? Not because of quality — because of three specific betrayals. I want to go through them one at a time, because each one cost us something before we understood it.

The first is the obvious one: generic presets are recognisable as generic. Your viewer has scrolled past that elastic bounce four hundred times this month. They could not describe it to you — nobody consciously catalogues text animations — but the pattern-matching part of the brain absolutely registers it. Seen this before becomes nothing new here becomes a flick of the thumb. The preset is doing its two-click job and quietly telling the audience this video is like the other ones.

There's a subtler cost too. When your face, your script, and your years of built-up trust appear on screen wrapped in someone else's visual language, the package contradicts itself. The person is specific; the presentation is stock. Viewers can't articulate the mismatch, but they feel it the way you feel a suit that doesn't fit — nothing you can point at, everything slightly off.

I felt this hardest when I started paying attention to whose content survived on mute. Most short-form video is watched with the sound off at least part of the time. On mute, your voice is gone, your energy is gone, your music is gone. What's left is footage and type. If the type is from a pack, you've just handed the muted version of your entire channel to a stranger's design decisions.

The second betrayal: motion that fights the edit

This one took me longer to see, and it's the one I now consider the real amateur tell.

Every edit has a rhythm. It comes from the speech — where the pauses land, where the emphasis falls — and from the music, and from the cuts themselves. A good editor is essentially a drummer. The cut lands on the beat, the b-roll enters on the phrase, the pause before the key line gets held exactly long enough to create the little vacuum the line drops into.

A template has its own rhythm, baked in by whoever made it. The bounce takes twenty frames because the pack designer decided it takes twenty frames. That designer never heard your client's voice. They don't know that she hits the key word early, or that the music you licensed has a push at the end of every second bar.

So the preset lands where it lands. Sometimes it accidentally agrees with the edit. Mostly it doesn't — the text settles a few frames after the word is spoken, the transition whooshes across a cut that wanted to be hard, the graphic is still animating in while the sentence has already moved on. Individually these are tiny. Accumulated across a video, they produce that feeling everyone knows and nobody can name: it just feels a bit off. A bit cheap. You can't point at the frame that's wrong because no single frame is wrong. The relationship between motion and sound is wrong, everywhere, slightly.

The fix is unglamorous. In our timelines, type is animated against the audio waveform. The editor scrubs the voice, finds the exact frame the stressed syllable lands on, and that is the frame the word arrives on screen. Not around there. There. It's slow work the first time and quick work once it's habit, and it is invisible when done right — which is the point. The viewer never thinks "well-synced typography." They think "this person is worth listening to," and they cannot tell you why.

Motion that lands on the audio's beat is invisible. Motion that misses it by four frames is why the video "feels cheap."

The third betrayal: when everything moves, nothing matters

The third betrayal is the one I'm least entitled to lecture about, because it took me years to grow out of it myself.

When you first get access to motion tools — or a pack full of presets — everything is exciting and everything gets animated. Every caption wiggles. Every noun gets an icon. Every cut gets a transition, every transition gets a whoosh. The video is busy, and busy feels like production value when you're new. I have exported videos like that with genuine pride. I'd show a cut to a friend and think the movement was the impressive part.

Here's what all that movement actually does: it removes your ability to emphasise anything.

Emphasis is relative. A word in bold only reads as bold on a page of regular text. If the whole page is bold, you don't have a bold page — you have a page with no emphasis at all, printed heavier. Motion works exactly the same way. If every caption bounces, the bounce means nothing. When the genuinely important number arrives — the claim the whole video was built to deliver — you have no volume knob left to turn. You already spent your loudness on the word "the."

When everything moves, nothing matters.

The hardest note I give editors — and I give it more than any other note — is remove half of it. Not because the animation is bad. Usually it's good; that's what makes the note hard to hear. But a timeline where six things are animated and one of them is the point is a timeline where the point is losing. New editors hear "remove half of it" as criticism of their effort. The experienced ones hear it as the actual craft: restraint isn't the absence of skill, it's the most expensive skill in the room. Anyone can add movement. Deciding what stays still — so that movement means something when it comes — is the job.

I say this as someone who did not want to hear it either. Some of my early cuts, if I dig them up, look like the timeline had a caffeine problem. Growth, for me, was measured in things I stopped doing.

What custom motion actually looks like

Fine — so if not templates, then what? "Custom motion graphics" sounds expensive and vague, so let me make it concrete. Here's what it means inside our studio, at the level of actual work.

It starts with designed typography, not defaulted typography. Before anyone animates anything, someone chooses a typeface — deliberately, for this client, considering how it reads at phone size, how it sits against their footage, whether it matches how they speak. A doctor who explains things calmly should not have the same type as a fitness creator who shouts. Then the behaviour gets designed: how does type enter? Does it cut in on the syllable, hard and confident? Does it track in letter by letter? Does it hold and then release? These are small decisions, but a channel is hundreds of videos, and small decisions repeated hundreds of times become a personality.

Then the type is animated on the audio's beat, the way I described above — waveform open, syllable found, frame matched. This is the part templates structurally cannot do, no matter how good the pack is.

Data gets made visual instead of merely spoken. If the script says one method takes twenty minutes and the other takes two hours, that's a comparison, and comparisons want geometry — two bars, two clocks, a gap you can see. If the script walks through a process, the process wants steps that build on screen as they're spoken. Watch any strong educational channel with the sound off and you'll still follow the argument. That's not an accident; someone translated the argument into pictures.

Transitions get designed to carry energy across cuts — motivated by something in the frame, a gesture, a movement, a shape, rather than dropped in from a folder called "transitions."

And all of it gets written down into a system. This might be the least glamorous and most valuable part. For each client we maintain what is essentially a small visual constitution: these typefaces, these colours, this is how type enters, this is how numbers appear, this is what we never do. It means video forty looks like it belongs to the same channel as video four. It means a new editor joining the account produces work that matches. Consistency is not a creative limitation — it's the mechanism by which an audience learns to recognise you.

The Shilpa test: recognisable on mute

Everything above sounds nice in theory. Here's what it looks like when it compounds.

When we started working with Dr. Shilpa Arora, she had around three thousand Instagram followers, and most of those were bots. Today she's at a million on Instagram and over seven hundred thousand subscribers on YouTube. I'm not going to claim motion graphics did that — the content did that, her knowledge did that, years of consistency did that. But one specific thing we built for her run is the cleanest example I have of what a motion language buys you.

Early on, we designed an on-screen typography system for her videos. Big, set titles — MORNING ROUTINE, WHAT I EAT IN A DAY — with a consistent typeface, consistent placement, consistent behaviour. The same visual voice carried through the Ghee Series, through Banana Smoothie, through everything. It wasn't flashy. Deliberately not flashy. It was hers.

And here's the test it eventually passed: you could take almost any frame from those videos, mute it, strip the context, and someone who followed her would know it was a Shilpa video. Not because her face was in every frame — often it wasn't. The type told you. The system told you.

Think about what that means mechanically, in the feed. A follower is scrolling with the sound off. A frame of set typography flashes past. Before any conscious thought happens, the pattern is matched: that's her. The thumb stops. That half-second of recognition, multiplied across every video and every follower and every scroll session for years — that's the compound interest of a motion system. A template pack cannot pay it, because the whole premise of a template is that ten thousand other channels are running the same visual language. You cannot rent recognition. It only accrues to things that are yours.

If a follower can recognise your video on mute, from one frame, your motion system is doing its job.

The objections, honestly

I run a studio that sells exactly this, so yes — obviously biased. But let me take the counterarguments seriously, because some of them are right.

Templates are cheaper and faster. True, and sometimes that's the correct trade. If you're just starting out, posting your first fifty videos, and the choice is between shipping with templates or not shipping at all — use the templates. Volume and reps matter more than identity when nobody knows you yet. We used them at that stage; I'd be a hypocrite to say you shouldn't. The problem isn't starting with templates. The problem is being three years and four hundred videos in, with a real audience, still wearing rented clothes.

Viewers don't notice motion graphics. Half right. They don't notice good ones — that's the definition of good ones. But they absolutely register sameness, and they register rhythm being off, even if the registration never becomes a thought. "I don't notice it" and "it doesn't affect me" are different claims, and only the first is true.

Custom is slow. At the start, yes. Designing a system takes real time — ours take weeks of iteration before they settle. But the economics flip after that, because a system is reusable. The slow part is building the language; speaking it is fast. By the twentieth video, a good custom system is often quicker to execute than hunting through a preset folder, because the decisions are already made.

You would say all this — you sell it. Fair. All I can offer is that the logic doesn't depend on hiring us, or hiring anyone. If you edit your own videos, you can do the core of this yourself this week: pick one typeface, decide one way that type enters, animate it on the syllable, and delete every other moving thing that has no job. That's a motion system. Version one of ours wasn't much more.

The right movement

I still see that bouncing title preset every single day. It doesn't bug me the way it used to. Now it just reads as a timestamp — it tells me a creator is in year one, doing what year one requires, the same way we did. No judgement. Everyone's first system is someone else's system.

But somewhere in a channel's life there's a quiet graduation: the moment its videos stop borrowing a visual language and start owning one. From the outside it's almost invisible. Nobody comments "great typography system." The videos just start feeling more like someone — more deliberate, more settled, easier to recognise, harder to scroll past. The motion disappears into the work, guiding and emphasising and bridging without ever asking to be admired.

Below this essay there's a piece we produced. If you watch it, watch it with this lens: notice what moves, when it moves, and what stands still so the moving thing can matter. Motion designed to guide, not decorate.

That's the whole aspiration, honestly. Not more movement. The right movement.

A With Media production — motion designed to guide, not decorate.