Cutroom
Use cases
Same take, different surface. What each channel actually rewards, and how to hit it.
One take in, *a finished TikTok ad out*
Upload one talking-head take of up to three minutes. Get back a finished 9:16 MP4 with captions burned in, ready for TikTok Ads Manager. A thirty second ad costs about 110 credits. Here is what goes in and what it costs, why TikTok buries imported creative, what the first second has to do, and how many openings a real test needs.
Targeting moved into the ad. *Creative is the audience now.*
Broad delivery means the system decides who sees what by reading the creative and the response to it. The ad has become the targeting lever. Running four of them is choosing four audiences. Here is how the skill moved, what the win rate says about sample size, how one take becomes a dozen audiences, and what it costs to kill a loser on Wednesday.
Your ad is judged against *the post from a friend*
A Reels ad is not competing with other advertising. It appears between somebody's holiday photos and a dog video, and it is compared to those. That is a harder standard than any category benchmark. Here is why polish works against you, where the interface eats your frame, how one take becomes a weekly rotation, and what that rotation costs.
They arrived to watch. *That buys you a longer sentence.*
A Shorts viewer opened an app whose purpose is watching, frequently with headphones in. Attention is a little more patient here than on a social feed. Here is how much that buys you, why a named speaker outperforms a brand voice, how to cut a read for this surface, and how one take supplies every vertical placement you run.
Sound is on and the screen is full. *The first word decides it.*
Snapchat plays an ad at full screen with audio running. There is no muted grace period to hide in. Whatever is said in the first second is the entire test. Here is why the audio is the hook, what this audience does to a persuasion structure, how to cut for a loud open, and why openings are tested in sets here.
The words explain it. *Your own footage shows it.*
Upload one talking-head take of up to three minutes and get a finished 9:16 MP4 with captions burned in, for about 110 credits. Your product footage sits underneath, on the words that describe it. Here is what goes in and what it costs, why demos should follow objections, how one product shoot serves ten demos, and what captions have to carry.
One unfakeable detail carries it. *The rest is decoration.*
A testimonial works on a single sentence containing something nobody would bother inventing. The fridge in the background, the exact week, the part they were wrong about. Polish removes those. Here is why specificity is the whole format, how to collect takes that contain it, why cutting one is deletion work, and what twenty customers cost.
You can say what a brand account cannot. *That is the format.*
A founder can admit the product is wrong for some people, explain a decision that looks odd, and name a competitor. A brand voice has none of those permissions. Here is why that is the mechanism, why the programme usually stalls at the edit, how to direct a take by deleting the hedging, and what twenty answers cost.
Interface footage converts *people who were already going to install*
A montage of screens works on warm traffic and does nothing cold. A stranger has no reason to study a menu yet. What moves them is the moment before somebody reached for the app store. Here is why situations beat interfaces, what the store listing is quietly costing you, how one take becomes many situations, and what to optimise instead of install cost.
Change one line, hold everything else. *Otherwise nothing is learned.*
Most hook tests change the opening, the pacing and the caption style together, then declare a winner. That result is unreadable. Here is why a three variable test teaches nothing, how many openings the arithmetic actually demands, which number judges a hook, and how one body holds identical across sixteen of them.
They already know who you are. *Skip the introduction.*
Most retargeting creative is the cold ad run again. It spends its opening explaining a brand the viewer already visited, at a higher cost per impression. Here is what that wastes, why one ad cannot cover every blocker, why a small pool makes frequency the enemy, and how each blocker becomes a new opening rather than a new shoot.
The side-by-side is restricted *and nobody believes it anyway*
Before-and-after imagery sits under platform restrictions in health, beauty and fitness. A decade of lighting tricks has taught viewers to discount it. The described version survives both problems. Here is why the two problems arrive together, how to describe a before in behaviour, why the after should be modest, and what a dozen befores cost.
A spoken explainer, *not an animated one*
Upload one talking-head take of up to three minutes and get a finished 9:16 MP4 with captions burned in for about 110 credits. It is a person explaining one idea, cut tightly. Here is what goes in and what it costs, why animation hides a broken script, why one idea per video wins, and why six short explainers beat one long one.
A promo that teaches beats *a promo that announces*
Upload one talking-head take of up to three minutes and get back a finished 9:16 MP4 with captions burned in for about 110 credits. Use it to teach one idea. Here is what goes in and what it costs, why a module list only converts your own list, which register loses a serious buyer, and what a launch set costs.
One take in. A finished ad out. See it on your own footage before you decide anything.