ClipRoute

Free clipping guides and courses: Clip Kit, Campaign Filter, Launch Protocol, Edit Standard, Retainer Route, Campaign Radar, AI Clipping...

Bacău, RO

•

•Created by

Cristian Chitimus

12 joined

Home

Products Reviews

Cristian Chitimus

@cristianchitimus·5d

Same channel, same hour, 5x the views. The only thing we changed was the topic.

Yesterday our second channel posted a Warhammer clip: 59 views in 24 hours. Today, same channel, same time slot, we posted a Sky: Children of the Light clip instead: 321 views in about 7 hours.

Why: that channel grew on animated shows. YouTube tests every new Short on the people who already watch you first. Show them something they don't care about, they swipe, and the feed stops pushing it.

What we do now:

check your channel's top 5 Shorts before picking a campaign — that's the audience you actually have;
one niche per channel; a new niche gets its own channel;
give a test clip 24 hours, and set the stop rule before you post (ours: under 100 views = switch).

Our main Sky channel also crossed 1,000 views on yesterday's clip, which is the campaign's payout threshold.

Cristian Chitimus

@cristianchitimus·Oct 8

3 payouts in, $14.15 received. Small, but the loop works end to end.

Our Content Rewards balance just hit $14.15 from 3 payouts, and another $16.44 is still accumulating on approved clips. Yesterday's Sky: Children of the Light Short passed 1,000 views in under a day.

How the money actually moves, since a lot of people ask:

Clip approved → it sits in Queued while it keeps collecting views.
The sponsor closes a payout cycle (no fixed date) → Arriving (short hold).
Hold clears → Received → your Whop balance → Withdraw.

What's working for us, one line each:

one campaign per channel, same hours every day;
submit 4–5 minutes after the clip goes public (not instantly, not after 30 minutes);
every title opens a question the clip answers in its last third;
don't post a new clip on a channel while yesterday's is still climbing — they compete for the same feed.

Not big money yet. But it's real, and it's growing.

Cristian Chitimus

@cristianchitimus·Oct 4

Before/after: the draft a brand rejected in one line, and the 10-second fix

Yesterday a brand declined one of our drafts with four words: "remove creator watermark".

BEFORE: we cut a clip from an approved source video. The edit was clean: hook in 2 seconds, the product fully visible, the brand name spelled right, our own captions. But in the first 3.7 seconds, the original creator's Instagram handle sat in the bottom-left corner of the footage. We never noticed it, because we were looking at our edit, not at the source.

AFTER: same clip, same cut points. On that one segment we crop to 87% of the frame, anchored at the top, so the bottom strip with the handle is gone and the subject stays whole. Resubmitted this morning.

Why it matters: when a campaign pre-approves drafts, every rejection costs you a day. The brand is paying for its own name on screen, not for somebody else's.

The check we added before any draft goes out:

1) Export one frame per second from EACH source segment, at full size.

2) Look at the four corners only: handles, channel logos, subscribe bugs, burned-in captions.

3) If something is there: crop that segment, or pick another take. Never blur it (a blurred box reads as "stolen").

Ten seconds per clip. Cheaper than a day in the pending queue.

Cristian Chitimus

@cristianchitimus·Oct 3

New effect: illustrations that tell a story, not one icon per sentence

Most talking-head edits illustrate sentence by sentence: "money" -> money icon, "time" -> clock. It looks busy and nobody remembers it.

We rebuilt a different system this week, from a reel made entirely with HyperFrames:

ONE hero object with a face that IS the number (an orange ball labelled "$10K").
- Every cost is a small monster in its own colour that bites the hero (18% interest = a red flame monster, 9.5% = a calmer purple blob, 0% = a sleeping green ghost).
- A counter under them changes on the exact word: "you pay $1,800/yr" -> $950 -> $0.
- One background colour per chapter, 2-3 words of text, the punchline word in serif italic.
- The finale: all monsters in a row with their yearly cost, the worst one crossed out.

Why it works: the viewer stays to see what happens to the hero. The illustration has an arc, so it carries retention instead of decorating it.

It's drawn from code (SVG + GSAP), so the style is identical on every clip and costs nothing per video. It's now a module in FXLIB (GX.HERO_COUNTER): you describe the beats in a JSON plan and the panel renders itself above the speaker.

Which line in your last video had a number that changed? That's where this goes.

Cristian Chitimus

@cristianchitimus·Oct 2

Test result: our "dead" channel wasn't dead. One clip was eating the others.

This week four of our Shorts on one channel stopped at 10-50 views. Same format, same campaign, same hours as clips that did 2,000+. The obvious read was "the channel got throttled". We tested it instead of guessing.

The test: we posted a clip from a different campaign (different topic, different audience) on our brand-new channel. Result: 357 views in 18 hours, while the campaign clips on that channel had done 5-13.

What actually happened on the old channel: one clip took off (13,500 views and still climbing) and every sister clip we uploaded after it, same series, same audience, same title prefix, got no test traffic at all. The feed was serving the winner to that audience, not the new ones. Not a penalty. Cannibalization.

The rule we run now:

While a clip is still gaining views every day, do not post another clip from the same series to the same channel.
- Wait until it stalls (2 days with no growth), or post the next one on a different channel.
- Rotate topics on a channel instead of stacking 8 near-identical clips in 6 days.

If your channel "died" right after your best clip ever, check what that clip is doing before you start a new account.

Cristian Chitimus

@cristianchitimus·Oct 1

Result: 3 of 4 clips passed 1,000 views on a brand campaign. The one that failed had the same scenes.

Quick result from a game campaign we run on YouTube Shorts (the "pass 1,000 views" kind of payout).

Four clips, same structure: a creator's face in frame 0, four scenes from the official trailer with one verified fact under each, the creator's reaction as the ending. Day-one views: 1,080 / 318 / 1,246 / (posted today). Same channel, same hours, same length.

The loser: a statement title ("Nobody wanted Van Gogh's art"), and the question on screen was answered in the first beat. The winners: a question title that the clip only answers in the last 3 seconds ("Why did he cut off his ear?", "Who made him famous?").

What we took from it: the question is the retention engine, not the facts. If the on-screen question closes early, the viewer leaves early. We now write the question first, then pick scenes that delay the answer.

Tooling note: all four were cut by an agent from a build script per clip (crop on the creator's face, fact captions timed to beats, two-pass loudness to -14 LUFS), verified on frame sheets before upload.

Cristian Chitimus

@cristianchitimus·Sep 30

We rebuilt a 21-preset text pack as code your AI can run (before/after)

Quick showcase from today's build.

A creator sent us a 3-second demo of a paid text-preset pack: 21 animated titles (highlight, shine sweep, strikethrough, blur-in, pop-up, metallic slide, drawn reveal, colour-cycling glow, 3D extrude, flicker, turbulence, typewriter, letter-by-letter, burning gradient, hand-drawn pill and the rest).

We rebuilt all 21 in FXLIB as JSON presets: real fonts, layered glow, 3D extrusion by stacked shadows, SVG draw-on strokes, GSAP entrances timed frame by frame to the reference (0-29 frames in, loop after). No After Effects, no timeline. An AI agent renders the whole grid from a 21-line plan.

How we checked it: reference vs ours, same frames, side by side, for every preset, then the full grid at 2x. Three rounds of corrections were about size and line spacing, not mechanics - measure on the final frame, not by eye.

What it means for you: any "preset pack" you like can become a parameter set your editor agent calls by name. The 21 IDs are TX.CC_* in the catalog.

Cristian Chitimus

@cristianchitimus·Sep 29

The 100-character rule that silently blocks your YouTube upload

Short tutorial from today's upload.

We had a Short ready, title written, description done, "not for kids" set, paid promotion set. Clicked Next three times. Nothing happened. No error, no red text.

The cause: the title was 108 characters. YouTube's limit is 100. Studio does not tell you. It just greys out Next.

Fix that took 20 seconds: cut the title to 91 characters and Next worked instantly.

Our rule now, before typing anything into Studio:

Count the title. Keep it under 100 characters, emoji included.
2. Put the hook first, the game or show name second, the campaign hashtag last. If it does not fit, the hook stays and the rest gets shorter.
3. If Next is greyed with no message, it is almost always the title length or a mention dropdown still open.

Small thing, but it cost a posting slot once. Now it never will.

Cristian Chitimus

@cristianchitimus·Sep 28

Before/after: where your clip starts decides if anyone stays

Same scene, same song, two different first frames.

BEFORE: the clip started where the lyric starts. Frame 0 = a black silhouette for 0.8 s (average brightness under 20/255). On a phone that reads as "video didn't load".

AFTER: we moved the start 1 second later, onto the source's own cut, where the character's face fills the frame. We lost one word of the lyric. Nobody misses it.

Why it matters: in our retention data, clips that open on a face + a question banner in frame 0 keep 50%+ of viewers to the end. Clips that open on anything else (black, a title card, an effect) sit under 40%.

The 20-second check we run on every clip before rendering:

Extract the first 15 frames and look at them as a grid.
2. If frame 0 is not a face, find the nearest cut in the source and start there.
3. Only then write the banner.

It costs nothing and it is the single biggest lever we have found.

Cristian Chitimus

@cristianchitimus·Sep 27

Hook before effects: what our retention data says

We compared our own Shorts on the same campaign (Prime Video UK) over the last week. Two groups:

A) A face and a question banner visible from frame 0. Nothing animated in the first 3 seconds.

B) An animated title card or a transition in the first 1-2 seconds, then the scene.

Group A: 50%+ of viewers still there at the end. Group B: under 40%, and the drop happens in the first 2 seconds, before the scene even starts.

Same footage, same length, same channel. The only difference was what the viewer saw in frame 0.

Rule we use now: nothing moves on top of the video before second 3. The face and the question do the hooking. Effects come after, on the punchline.

If you use effects (FXLIB or anything else), put them AFTER the hook, not instead of it. Got retention numbers that say otherwise? Reply below.