What repurposes well is the idea, not the finished file. The video can be the same one, but the aspect ratio, the length, the hook in the first seconds and the caption get rebuilt for each network. An honest shoot yields about five pieces, spread over several weeks.
The four layers of a piece, and the one that crosses
A finished post has four layers and only one of them moves from one network to another untouched. The idea, the fact you are telling or the thing you are showing, travels whole. The hook, the format, the caption and the call to action are stitched to the network they were born on, and they are exactly what breaks in transit.
It is worth seeing how different they really are. On TikTok the hook occupies the first second and arrives with sound, so a video that opens with two seconds of logo has already lost. In an Instagram reel the cover and the first frame decide, because a lot of people watch on mute and from the profile grid. On LinkedIn what gets read is the two lines before the "see more" link and the video starts silent, so an argument that only exists in the audio does not exist.
It keeps changing as you go down the list. On Pinterest the piece is not consumed the day it goes out, it is searched for months later, which makes the title matter more than the hook. On a YouTube Short the title is a search term rather than a line of script. On Threads and Bluesky the unit is one short sentence, 500 and 300 characters, with a conversation behind it. On Google Business there is no hook at all: there is an update from a business that somebody is looking at on a map while deciding whether to walk in.
The practical consequence is that the work is not in moving the file, it is in rewriting the three layers that do not travel. If the shoot cost you two hours and adapting the pieces costs you 20 minutes, you almost certainly have not adapted them.
The file itself, and the traps that give a repost away
There is a second layer of problems that is not about the message but about the file, and it is the one that makes a viewer notice this came from somewhere else. These are the real limits each connected network accepts for a video, which is the piece that travels most.
| Network | Format | Aspect ratio accepted | Maximum length |
|---|---|---|---|
| Reel | 9:16 | 90 seconds | |
| Story | 9:16 | 60 seconds | |
| Feed | Between 4:5 and 1.91:1 | 60 minutes | |
| Reel | Vertical, 9:16 preferred | 60 seconds | |
| Story | 9:16 | 120 seconds | |
| TikTok | Video | 9:16 | 10 minutes |
| YouTube | Short | Vertical required | 180 seconds |
| YouTube | Video | 16:9 | No documented limit |
| Feed | Between 1:2.4 and 2.4:1 | 30 minutes | |
| Threads | Feed | 9:16, 1:1 or 16:9 | 5 minutes |
| Feed | 2:3, 1:1 or 9:16 | 15 minutes | |
| Bluesky | Feed | 16:9 or 1:1 | 60 seconds |
| Snapchat | Feed | Vertical, 9:16 preferred | 60 seconds |
| Google Business | Feed | No video accepted | No video accepted |
The four networks missing from the fourteen, Reddit, Telegram, Discord and WhatsApp, publish no video requirements worth checking in advance. And LinkedIn's 30 minute ceiling is the company page one: a personal profile stops at 10, and there is no way to tell the two apart from outside, so it pays to stay short.
What the table shows at a glance is that an 80 second cut fits an Instagram reel, does not fit a story, and sits comfortably inside TikTok. That is why the master piece is best cut at 9:16 and under 60 seconds: everything travels upward from there and nothing travels back. A landscape shot cropped to vertical at the last minute always shows, because the captions fall outside the frame and the head ends up glued to the top edge.
The floors matter as much as the ceilings and get forgotten more often. An Instagram or Facebook reel wants at least three seconds, Pinterest four and Snapchat five, so that brilliant two second cut has no destination on any of the four. Weight bites too: Bluesky recommends staying under 1 MB per image and 50 MB per video, which is between four and eighty times less than any other network on the list.
How many pieces one session really yields
The optimistic arithmetic says an hour of filming produces fifteen posts. It does produce that number, but half of them are the same sentence against a different background and the audience watching them is the same audience every time, so you have not multiplied reach, you have split the same reach into more parts and lowered the average of each one.
The split that holds up is shorter and more uneven. From a well prepared hour you get one full vertical video, two cuts under a minute that stand on their own, a carousel with what was said written down, and one long text for the places where text works. Five things. Each answers a different question, and that is the proof they are five pieces rather than one piece repeated five times.
For that split to exist you have to decide it before you film. In practice that is three things: leaving clean silences between blocks so you can cut, filming some b roll that can serve as the carousel cover, and noting down while you film the two or three sentences that will become the long text. Anyone who tries afterwards gets two and a half pieces.
What actually multiplies is not the number of parts, it is the number of networks each part can visit. Five pieces across four accounts is 20 appearances without anybody seeing the same thing twice, as long as each caption is written for the place it appears in.
The order they go out in, and the gap between cousins
The order is not cosmetic. The piece that introduces the idea should go out where the native format is strongest and where the algorithm distributes fastest, which today is the short vertical video. The long version and the written version come afterwards, because they ride the interest the first one raised and because they reach people who do not consume video.
Between a piece and its direct cousin, the one cut from the same material, leave at least a week when they go to different networks. If they go to the same network, the distance rises to a month and the piece has to change format, not just caption. More of your audience follows you in two places than you think, and the cost of being seen twice in two days is not that they get bored: it is that they stop opening your notifications.
That calendar runs into the other problem, which is how much you publish in total. Pushing five derivatives of one session out in a single week raises your apparent frequency without raising your production, and the usual result is worse average engagement across all of them. How to work out the number you can hold without diluting yourself is in the posting frequency guide, and it is worth reading before you decide how many derivatives to use.
Everything in this article happens in one place in GoFeed.
Try it freeWhat gets rewritten every time
The caption comes first, and it is where copy and paste shows from a mile away. An Instagram caption, which can run to 2,200 characters, opens with a sentence that hooks and leaves the hashtags at the end or in the first comment. That same text on LinkedIn reads like an advert, because there the first two lines are expected to frame a professional problem and nobody wants a stack of hashtags. On Threads it does not fit at all: 500 characters. On Bluesky, 300, so you keep the sentence and drop the paragraph. And on Pinterest, another 500, but written the way a headline is written, with the words somebody would search for six months from now.
The call to action comes second. "Link in bio" works on Instagram and means nothing on a YouTube video, where the link lives in the description, or on Google Business, where the button is whichever one the platform offers. "Comment the word guide and I'll send it" makes sense where there is a direct message flow behind it and is noise where there is not. Changing the call to action costs one line and is the difference between a piece that works and a piece that decorates.
The third is the first line of the video. You rarely need to refilm: trimming two seconds off the front, putting a different caption on screen and changing the cover is usually enough for the same recording to stop looking like a forward on the second network. And if you have several accounts connected, the place to write those variants side by side, with each network's limit visible while you type, is the publishing calendar.
The part that can be a standing rule
Half of all this is craft and does not automate. The other half is downloading a file from one account and uploading it again to another, and that half can stop being done by hand. A Workflows rule joins one origin account to one destination account, one to one, and either carries the piece across as soon as it appears or drips it into the time slots you set.
Three things to know before you build one. The origin can be Instagram, Facebook or a LinkedIn company page, and nothing else: TikTok, YouTube and Google Business will answer a question about a past post with the record but not with the video file, and without a file there is nothing to carry. The drip distributes what the workspace has collected since you switched the rule on, not the whole back catalogue. And the caption is copied as it stands, or gets a prefix and a suffix you typed yourself, or is left empty; nothing is written for you.
It is not a recycling queue either. It does not repeat the same piece on the same network over and over, and a database constraint makes a second delivery of the same source post impossible. What it does is carry a piece once to a different network. Bringing back something that worked three months later is still your work, rewritten and with a different hook, which is how it should be.
Which pairs of accounts can be joined, what it warns about and what it refuses is covered in cross posting automation, and the comparison against the tools that only do this one slice is in GoFeed vs Repurpose.io. The whole module lives at Workflows, and its only promise is that one: the transport is automatic, the writing is yours.