How to Turn a YouTube Video Into a Newsletter

Quick answer: To turn a YouTube video into a newsletter, get the transcript, then rebuild it as an argument rather than tidying it line by line: pick the one idea worth a full issue, restructure the spoken material around it, keep two or three direct quotes, and add what only you can say. Use the leftover moments for Notes or social posts. Narrareach fetches transcripts from YouTube, podcast feeds, Apple Podcasts and direct audio links, and your AI assistant drafts the article and Notes from them in your voice.

A recorded conversation is often the best thinking a writer does all month. Someone asks a sharp question and you answer it better than you would have alone. Then the recording goes up, a small share of your readers press play, and the best line sits at minute 27 where almost nobody will hear it.

The fix is not to "repurpose" the video into a summary. It is to write the issue the conversation was pointing at. This guide walks through how, from getting a clean transcript to the draft, plus the rights question most people skip.

Why turn a video or podcast into a newsletter?

Most newsletter readers prefer reading. They open your email because they can get through it in five minutes, on a train, without headphones. A link to a 45-minute video asks them to change how they consume you, and most will not.

Writing also forces precision that speaking does not. In a conversation you can circle an idea three times before landing it. In an issue you have to land it once. That rewrite is where the recording becomes something people forward.

And it is efficient. The research, examples and opinions already exist on tape. Nothing starts from a blank page.

Step 1: Get a clean transcript

For a YouTube video, you need the link to the video itself. A channel or playlist link will not work. YouTube's own transcript panel gives you the text, but it comes with timestamps and line breaks every few words, and you have to clean it before you can write from it.

Podcasts are harder. Many hosts do not publish a transcript, so you need the audio. The RSS feed is almost always public, and it links straight to each episode's audio file. Apple Podcasts links work the same way. Spotify is the usual dead end, because it protects its episodes. If all you have is a Spotify link, search for the show in Apple Podcasts or find its RSS feed instead.

In Narrareach, you paste any of these into a conversation with your AI assistant (Claude or ChatGPT, connected through the Narrareach MCP connector) and ask for the transcript:

• YouTube videos come back in seconds.
• Podcast RSS feeds and Apple Podcasts links use the latest episode unless you name one by its title.
• Direct audio links (https only) work up to three hours. MP3 and WAV are split into ten-minute pieces, so length is the only limit. M4A, Ogg, WebM and FLAC are sent in one piece and need to stay under about 24 MB. An MP3 copy of a large file goes through.
• Videos you uploaded to Narrareach stay private to you.

Long audio is transcribed in the background. Your assistant gets an id and checks back until the text is ready, which can take a few minutes for a long episode. By default the transcript returns up to 60,000 characters. Ask for more, up to 400,000, if you need the whole of a long interview.

To show what comes back: I tested the direct audio path with the LibriVox reading of the Gettysburg Address, a public-domain recording. The audio is 158 seconds long, and the transcript came back as 1,660 characters of plain, punctuated text, starting "Four score and seven years ago, our fathers brought forth on this continent a new nation..." No timestamps to strip, no line breaks every few words.

Step 2: Find the one idea worth a full issue

Do not try to cover the whole conversation. A 40-minute talk usually holds one idea that deserves 1,200 words and five or six that deserve a paragraph each.

Read the transcript once quickly and mark three things:

1. The moment the energy changed. Where the speaker leaned in, disagreed or told a story. That is usually your thesis.
2. Lines you would quote. Sentences that work on their own, out of context.
3. Loose threads. Ideas raised and dropped. These become Notes and follow-ups, not part of this issue.

Then write one sentence that states the issue's argument. If you cannot, read the marked sections again. Everything that does not support that sentence moves to the leftovers pile.

Step 3: Restructure, do not transcribe

Spoken language is repetitive and nonlinear, and it should be. A newsletter is not. The most common mistake I see is an issue that reads like a cleaned-up transcript: technically accurate and hard to finish.

A structure that works for most conversations:

• Open with the sharpest moment, even if it came late in the recording.
• State the argument in your own words.
• Support it with two or three sections, each built around one example or exchange from the conversation.
• Quote the speaker directly two or three times, so their voice sits next to yours.
• Close with your view: where you agree, where you do not, and what the reader should take away.
• Link to the recording at the end for readers who want the full context.

With Narrareach, once the transcript is in the conversation you ask your assistant to write the article. It reads your voice profile first, writes from what the transcript actually says, and saves the result as a draft in your Narrareach Drafts. Give it the brief you would give a human editor: which section to build around, who the issue is for, what to leave out. Nothing is scheduled at this stage. You open the draft, add the parts only you can write, and schedule it to Substack when it is ready.

Step 4: Check every quote against the transcript

This is the step that protects your credibility. Whether you or an AI wrote the draft, search the transcript for every sentence in quotation marks. If it is not there word for word, either fix it or turn it into a paraphrase without quotation marks.

The assistant connected to Narrareach is told to quote only what the transcript says, but the check is still yours. It takes five minutes and saves an awkward correction email.

Step 5: Turn the leftovers into Notes

Those loose threads you marked are a week of short posts. Each one carries a single idea from the conversation, in a few sentences, with a link to the full issue once it is live.

In Narrareach, your assistant can pass the same transcript to the Notes generator. It writes up to 10 Notes per request in your voice, each with a suggested posting time, for Substack, LinkedIn, X, Bluesky, Threads or Facebook. Generation saves and schedules nothing. You keep the ones that sound like you, and only those go into the queue. It is the same step as repurposing an article as Substack Notes, starting from audio instead of text.

Do you have the right to turn someone else's video into a newsletter?

Not automatically. Transcribing is easy; reusing is a rights question. Safe sources are your own recordings, guest appearances where you agreed the material could be reused, and content whose creator gave you permission.

Quoting a short passage with credit and a link, inside your own commentary, is ordinary practice. Publishing someone else's talk as your article is not. Narrareach does not check who owns a video or episode, so that judgement stays with you.

What does it cost in time and money?

By hand, the slow parts are getting usable text and writing the first draft. For a podcast without a published transcript, you also need a transcription service.

In Narrareach, transcripts are part of Revenue Writer ($39/month) and Full Agentic Mode ($89/month), and the 7-day trial. Revenue Writer includes 100 YouTube transcripts and 120 audio minutes a day. Full Agentic Mode includes 300 and 600. A public video or episode that someone has already transcribed comes from a shared cache and does not count against your allowance. On Full Agentic Mode you can also automate it through the REST API or the n8n node, if you want every new episode transcribed without asking.

Frequently asked questions

Can I turn a YouTube video into a Substack newsletter?

Yes. Get the transcript, restructure it around one argument, and publish it as an issue with a link to the video. Narrareach fetches the transcript and drafts the article in your voice, saved to your Drafts for you to approve and schedule.

How do I get a transcript of a podcast episode?

Use the show's RSS feed, an Apple Podcasts link or a direct link to the audio file. Spotify links cannot be transcribed, so use one of the others for the same episode.

How long can the recording be?

In Narrareach, up to three hours of audio. MP3 and WAV work at any file size up to that length. Other formats need to stay under about 24 MB.

Should the newsletter cover the whole video?

No. Pick the one idea that deserves a full issue and turn the rest into Notes or short posts. A summary of everything is harder to read than the video was to watch.

Will Narrareach publish the article automatically?

No. The article is saved as a draft and the Notes are returned for review. Nothing is scheduled until you approve it.

Write the issue the conversation was pointing at

A good recording already did the hard thinking. The newsletter is where you finish it: one argument, the best quotes, your view, and a link back for anyone who wants the full context. If you have a recording sitting unwritten, start with that one.