An editor that learns how you edit
Dollop learns your editing style from videos you’ve already made. Add your published videos and a few creators you admire, and it measures pacing, hooks, captions, B-roll, framing, audio and color. The result is a style profile written as plain-language rules, each with the evidence behind it, and every first cut and chat edit follows it.
What it measures
Each reference video is analyzed for the things that make an edit feel like yours. A finding is only recorded for what was actually measured, and each one points back to the moments in the videos it came from.
- Pacing: cuts per minute, shot lengths, pauses and words per minute
- Hooks: how long the opening runs, how many cuts land in the first five seconds, and how it starts
- Structure: sections, calls to action, sponsor disclosures and chapters
- Shots: talking head versus B-roll, how large your face sits in the frame, how long cutaways run
- Captions: whether they’re on screen, where they sit and how many words show at once
- Audio and color: loudness, music under speech, brightness, saturation and warmth
Where references come from
Paste links to YouTube, TikTok or Instagram videos, or upload the files. You can also paste a channel or profile link and choose which of its videos to learn from. When a link can’t be downloaded for full analysis, Dollop says how much it could measure instead of guessing.
Mark each reference as your own published video, a raw and finished pair, inspiration from another creator, or a brand reference. A raw and finished pair teaches the most: Dollop lines the two up and sees what you kept, cut and reordered. From other creators it learns only abstract editing traits, never their clips, branding, voice or music.
Rules you can read and change
Your style profile is a list of rules grouped by topic, such as “remove filler words”, “no transitions” or “word-highlight captions, four words per line”. New findings arrive as a proposal that shows what would change, and nothing changes until you approve it.
Your own rules beat learned ones. You can add rules for what never to do, lock the profile so learning can’t change it, and go back to an earlier version from its history. Safety rules, such as keeping skin and product colors as filmed, always stay on.
Profiles that stack
Your main style is the baseline every video starts from. Platform and format styles build on it, so your TikToks can move faster than your long-form reviews without starting over. The more specific profile wins, and safety rules apply everywhere.
It keeps learning from your edits
When you change the AI’s cuts, Dollop looks for patterns in what you changed and suggests them as rules. Keep the ones you want and dismiss the rest.
Questions
No. From inspiration references it learns only abstract editing traits, such as pacing and caption placement. It never reuses their clips, branding, voice or music.
No. Footage minutes count the video and audio you upload to projects. Style references don’t count.
A studio keeps up to 20 references at a time, and each reference video can be up to 10 minutes and 500 MB. Delete ones you no longer use to make room.
Yes. Each finding points to the moments in the reference videos it was measured from, and a proposal shows what would change before you approve it.
Related features
AI first cut
Turns raw footage into an editable first cut: retakes, filler words and long pauses out; disclosures and product mentions in.
Learn moreEdit by chat
Ask for a change in plain words and review it as a scoped edit before it’s applied.
Learn moreAuto captions
Captions from your transcript in five styles and ten fonts, placed inside each platform’s safe area and exported as previewed.
Learn more