TLDR
A published video is the most under-used asset in most content operations: the hard thinking is done, but the ideas are locked in a format only one platform rewards. Mine it — pull the transcript, extract every self-contained idea (claims, frameworks, numbers, quotable lines), and turn each one into a designed post. Done by hand it takes an afternoon per video; an AI carousel generator that accepts the YouTube link directly compresses the extract-and-design part to about a minute per post.
Why repurposing beats another new post
The most expensive part of content is the thinking: the argument, the framework, the examples, the numbers. When you publish a 30-minute video, that expensive part is done — and then most creators walk away from it and start the next idea from a blank page.
That's backwards for two reasons. First, the yield: a substantive 30-minute video contains far more usable material than one post can carry — a handful of distinct claims, at least one step-by-step framework, and a few lines worth quoting verbatim. Second, the audiences barely overlap. YouTube rewards watch time; LinkedIn and Instagram reward swipeable, scannable ideas; X rewards compressed takes. The people who would love your argument on LinkedIn will mostly never see the video it came from. Repurposing isn't recycling — it's distribution the algorithm silos otherwise deny you.
The catch, and the reason everyone nods at this advice and few follow it: extraction is tedious. Rewatching your own 30-minute video with a notepad feels like homework. The rest of this guide is about making that part cheap.
What one video actually contains
Repurposing goes wrong when you try to summarize the whole video into one post. A summary flattens everything interesting; the value is in the individual ideas, each strong enough to stand alone. Work through the video hunting for five specific things:
- Core claims — the statements the video exists to make. Each one is a carousel hook or a thread opener.
- Frameworks and step sequences — "the 4 stages of…", "how I do X". These are the highest-value finds: a numbered framework maps one-to-one onto carousel slides.
- Numbers and results — concrete figures, before/afters, timelines. Specifics carry posts; vague lessons don't.
- Quotable lines — sentences you said better out loud than you'd write them. These become quote graphics with near-zero extra work.
- Objections you answered — every "you might be thinking…" moment is a ready-made post for the skeptical half of your audience.
A 30-minute video that yields three claims, one framework, four numbers, and two quotable lines just became 5–10 posts. That's the inventory; now it needs packaging.
The manual workflow
The honest version, if you're doing this by hand:
1. Read the transcript, don't rewatch the video
Rewatching is the step that kills the habit — 30 minutes minimum, plus pausing. Open the transcript instead — look for "Show transcript" in the video description; it's there whenever the video has captions, which auto-captioning provides for most uploads — and skim it in five minutes, highlighting the moments that match the checklist above. The transcript is ugly reading, but you're mining, not proofreading.
2. Pick the 3–5 ideas that stand alone
An idea qualifies if someone who never saw the video can get value from it in isolation. "Why most carousels lose readers on slide 2" stands alone; "the thing I mentioned earlier about hooks" doesn't. Cut ideas that need the video's full context — they make weak posts.
3. Rewrite each idea for the platform, not from the transcript
Spoken sentences don't work as slide copy — they're longer, looser, and full of connective tissue. For each idea, write a hook (the claim, sharpened), 3–6 supporting points (one per slide), and a closing beat. This is real writing, not copy-paste, and it's where most of the remaining manual effort lives.
4. Design the slides once, export per platform
Lay the idea out as a carousel — one point per slide, big type, consistent palette — and export at each platform's native size rather than stretching one image everywhere. If you need the numbers, our Instagram carousel ideas guide covers what makes slide sequences hold attention, and formats differ enough between platforms that even a carousel on X (Twitter) is its own small puzzle.
Total cost per video, done well: an afternoon. That's the price that makes "I'll repurpose more" the most broken promise in content marketing.
Which format each idea deserves
Not every extracted idea should become the same kind of post. Match the format to the idea's shape:
- Frameworks and step sequences → carousels. A numbered framework is already a slide sequence: one step per slide, the framework's name as the hook. This is the highest-conversion pairing in repurposing, and it's why carousels sit at the center of the workflow rather than being one option among many — swipeable posts are the format LinkedIn and Instagram feeds structurally reward.
- Core claims → carousels or single-image posts. A claim with 3–6 supporting points is a carousel; a claim that lands in one sentence is a bold single graphic. Don't pad a one-liner into five slides.
- Quotable lines → quote graphics. Your face or the line in big type, video title as attribution. Cheapest post per unit of effort in the whole system.
- Numbers and results → the hook slide of something bigger. A concrete number is the strongest first slide a carousel can have; spend it there instead of burying it mid-sequence.
- Objections you answered → threads. Objection-and-response has a natural back-and-forth rhythm that reads better as a text sequence than as designed slides.
One deliberate omission: don't turn the video into "watch my video" posts. A link with a thumbnail asks the feed's users to leave the feed, and every platform's distribution quietly punishes it. Repurposed posts carry the value themselves — the video becomes the deep-dive destination for the fraction who want more, mentioned in a comment or profile, not the post's job.
Where AI actually helps
"AI content repurposing" tools attack different slices of this workflow. Video-to-video tools cut long recordings into vertical clips — useful, but the output is still video on video platforms. The gap that matters for reach is video → designed slide content, because carousels and swipeable posts are what LinkedIn and Instagram actually distribute — and that's the slice an AI carousel maker covers.
Here's what that looks like with xcarousel, stated plainly: paste the YouTube link — no downloading, cutting, or transcribing; the AI processes the video itself, long-form included — and it extracts the key points and generates a designed carousel: copy split across slides, palette, typography, layout. The result lands in a real canvas editor, so the step-3 rewriting above becomes editing — tightening a hook, swapping a point — rather than writing from scratch. Export at native sizes for Instagram, LinkedIn, or TikTok, and publish to LinkedIn directly from the app.
What the AI doesn't do is choose which ideas deserve posts — it surfaces the video's main points, but you know which claim is the contrarian one worth leading with. The judgment stays yours; the tedium doesn't.
One video in, a week of posts out
The manual path works: transcript, checklist, 3–5 standalone ideas, per-platform rewrites, slide design, native-size exports. Budget an afternoon per video and it compounds nicely.
Or run the short version: paste the YouTube link into xcarousel, get a designed draft of the key points in about a minute, spend your attention on the two decisions that matter — which ideas lead, and how the hooks read — and repeat until the video is mined out. The 30-minute video you published last month is still sitting there, full of posts. Go get them.