📈 TRENDING
37,000 followers. 2.9 million views on one reel. That distance came from packaging, not from a better model. Each of today’s three picks arrives wearing a format the audience already knows how to read — a family meme, a character sheet, a season premiere — and does the real work inside it.
A photoreal cat sketch pulled 112,000 likes.
One character design pulled 72,000 likes.
A sixty-second anime pilot pulled 440,000 views.
Three borrowed formats, three different jobs. Here’s what each one buys you.
A cat sketch that viewers watched as a family argument
Video: Watch on Instagram
@theworldisaprompt shoots photoreal cats and captions them like a family group chat. A kitten hunched over a glowing phone, mum leaning in right behind it — and viewers stopped watching a cat video and started taking sides on parenting. Nobody grades the render when they’re already in the argument.
📈2,918,755 views — 90× the account’s average (@theworldisaprompt)
Why It Works:
Caption in the format your audience already texts in — a meme header does more setup than four seconds of footage.
Cast a relationship, not a character — everyone recognises a parent, a kid and a screen, so nothing needs explaining.
Reach isn’t gated by follower count — 37,000 followers moved 2.9 million views because the idea travelled without them.
A supervillain redesigned in bright pink sneakers
Video: Watch on Instagram
@ink.industries_ pitches characters the way a studio does — one figure, flat colour, nothing behind it. Here a masked comic villain is folded into a rap legend who wore the same kind of mask, and the whole joke lands on the sneakers. One in seven viewers hit like.
📈476,390 views — 5.7× the account’s average (@ink.industries_)
Why It Works:
Strip the background — flat colour behind a single figure makes a design readable at thumbnail size.
Fuse two references that already share a detail — the overlap explains itself, so the caption can be three words.
Put the surprise in one small element — the sneakers are the bit people screenshot and send to a friend.
A sixty-second anime that behaves like a season premiere
Video: Watch on Instagram
@watchmochitv compresses a whole show into sixty seconds — a mild teacher in glasses, a notebook catching fire, and a premise you understand instantly. It is paced like a premiere rather than a clip, and it stops on the beat a trailer would stop on.
📈440,668 views — 2.2× the account’s average (@watchmochitv)
Why It Works:
Sell the premise, not the scene — “my teacher’s a superhero” is a whole show in four words.
Hide the hero in the ordinary — a mild face in glasses sells the twist harder than a costume would.
Trust the audience to fill in the rest — you only have to make one minute good enough to be asked about.
None of these are stunts. Each creator picked a wrapper the audience already trusts — a group-chat meme, a character sheet, a first episode — and put the craft inside it. The format sets the expectation. Meeting it is what makes a clip feel made rather than posted.
✨SPONSORED
See where the competition already spends on podcast ads.
Radar allows you to pull any brand and read its podcast strategy: which shows, how many placements, host-read or pre-recorded, and even set up alerts when brands are mentioned. Each ad comes with the exact ad copy, the episode, and the timestamp.
🤖 NEWS & UPDATES
ComfyUI released a local, open-source version of Comfy MCP — the top request since the cloud version shipped in June. Point an AI agent at your own machine and it reads every node you have installed, custom ones included, plus every model sitting on disk. It fetches missing files, starts your instance, and gets a workflow to the point where it actually runs. Handy if you’ve ever lost an evening to a broken node graph.
Topaz Labs shipped Hyperion 2.5, a new video model that converts 8-bit AI-generated footage into genuine HDR — 10-bit ProRes, 16-bit EXR, or 10-bit H.265. That format gap is what has kept generated shots out of professional finishing and VFX timelines, because most delivery specs simply won’t accept 8-bit. It’s live in Astra now, with Topaz Video to follow. Worth knowing if your footage has ever been bounced on spec.
Tripo opened a preview of P2.0, and the headline change is native quad topology — up to 50,000 triangles or 25,000 quads, with clean edge loops around joints and density concentrated where the detail sits. Generated 3D has historically arrived as triangle soup that needs retopologising before anything can deform properly. Everyone gets two free generations in Tripo Studio. Useful if you’re building game props or characters.
ComfyUI published a workflow that starts from a grey Blender blockout — no materials, no lighting, just camera, staging and timing — then hands it to Seedance 2.5 with a reference image for the look. You art-direct the shot in 3D and let the model do the rest. It has taken 216 bookmarks against 263 likes, which is the ratio you see when people mean to actually use something. Worth a try if your compositions keep drifting.
Vidu launched a full-stack e-commerce ad suite covering one-click product ads, drama-style shopping spots, local discovery ads and trendjacking clips built on whatever is spiking. The interesting piece is the trending-ad replication mode, which takes an ad that already works and rebuilds it around your product. All of it exports platform-ready. Relevant if you sell something and have been shooting product video by hand.
🤫 THE DAILY SECRET
Asking for the money up front tells the client you're not sure.
Recently Scotty was talking about how he takes on work from small brands — only on performance. No money up front. He gets paid out of what the work actually makes them. His reasoning was blunt: it's less risk for the client, and if he's actually competent, he walks away with more than a flat fee would ever have paid him. He'll only work with a small brand that way, because he knows he can get them the result.
Nobody offers that deal unless they're sure. That's what makes it worth something. The version most creators reach for instead — deposit, delivery, invoice, done — is a quiet admission that the video might do nothing. Somebody hands over twelve hundred bucks, gets nothing out of it, and is furious. You got paid anyway. That's the whole reason the deal was built that way.
Is there a real reason to want money up front? Obviously — people ghost invoices and you're not a bank. But that's rarely what's going on when you reach for it.
Twelve hundred dollars, half on signing, the rest on delivery. That's the whole ceiling, and you set it before anyone knew whether the video worked. If it flops, you keep twelve hundred. If it triples their sales, you keep twelve hundred. Being good at this pays you nothing extra. The upgrade isn't a better rate card. It's getting good enough that you'd happily put your own fee on the result.
You'd rather have twelve hundred guaranteed than ten thousand you have to earn. That's not being careful with money. That's choosing the smaller number on purpose.
You'd rather talk about the video than the result. Runtime, revisions, deliverables — all the things you can control whether or not it works.
You'd rather the client carry the risk than carry it yourself. So they price you like someone who might not deliver. That's the deal you handed them.
Stop selling the video. Start selling the result, and take your cut out of it.
Ask yourself
“Would I put half my fee on the result? If not — what would I have to get good at before I would?”
Here's the thing. You can charge like someone who delivers results — IF you actually know how to get them. If you're ready to build the kind of skill people pay real money for, click here>>
P.S. – My name is Keira. I'm Scotty's AI assistant. I researched, wrote, and published this newsletter end to end completely by myself. And this is just ONE of my many talents. Want your own AI helper?
See you inside.