How Do I Repurpose a YouTube Video Into a Blog Post? 7 Steps

You published a video on Tuesday, and by Friday the lesson is still sitting on YouTube with no blog post. We'll set up a seven-step routine to repurpose a YouTube video into a blog post using your video's captions, an AI assistant, your blog editor and a spreadsheet. Setup takes an afternoon.

How do you repurpose a YouTube video into a blog post?

Here's the routine in one line: pick the one question the video answers, export the transcript, have an AI assistant draft the post from that transcript only, check the draft against the video, then publish. Once it's running, your part is picking the question, marking the tangents and doing the review.

It's built for creators and coaches who sell courses or programs and record more than they write.

You'll find more routines like this in our Playbooks library. The creators and coaches hub collects the ones for people who sell programs.

What do you need before you start?

Five things:

• A published or unlisted video with captions turned on
• A new, empty draft post in your blog editor
• An AI assistant
• A spreadsheet for the log
• One named reviewer: you, or a VA who can compare text with a recording. Name them now, because a check nobody owns doesn't happen.

Keep to one question per post. If a video drifts into a second topic, save that for another post.

What are the 7 steps to turn one video into one post?

Step 1: Pick the question and get the transcript

• Write the one question the video answers, the way a viewer would type it into Google.
• Add a row to your log with the video link, the question and the reviewer's name.
• Open YouTube Studio > Subtitles and click the video. On the caption track's row, click Options (three dots) > Download > .srt, so each line keeps its timestamp. Don't copy the plain text, because Step 5 needs the times.
• Test: the question would make sense typed into Google with no context.

Step 2: Mark the sections that drift off the question

Mark whole off-topic sections so the AI skips them, and leave the filler words. The AI usually drops those when it drafts, and you'll catch any leftovers when you read the page in Step 6.

• While you scrub the video, write the start and end time of each tangent or second topic in the video's row of your log, like "6:10 to 9:45".
• Test: every part you didn't mark helps answer the question.

Step 3: Make a names list

Auto-captions mishear names and program titles, and the AI copies those mistakes word for word.

• Copy every program title, brand and tool name from your sales page into a short names list in your log.
• Reuse the list for every video, and add new names as you launch them.
• Test: every name on the list is spelled the way your sales page spells it.

Step 4: Run the draft prompt

Tell the AI to use only the transcript. Without that rule, an AI assistant can fill the gaps with claims that sound right but that you didn't make. It will still slip sometimes, which is what Step 5 catches.

Draft a blog post using only the transcript below. Don't add facts, numbers, names or results that aren't in it. Keep each fact with the person who said or achieved it. If I describe a client's or guest's result, write it as theirs, never in first person. After each number, name or promise, put its transcript timestamp in brackets, like [06:12]. Spell these names exactly: [names list]. Tag anything you're unsure of with [CHECK]. Answer this one question in the first two sentences: [question]. Skip everything between these times: [time ranges]. Transcript: [paste]

• Replace [question] with your question, [time ranges] with the ranges from your log and [names list] with your names list, then paste the caption file where it says [paste].
• Paste the result into your draft post.
• Test: pick three numbers or names in the draft and find each one in the transcript.

Step 5: Check the draft against the video

Here's the part nobody tells you: the line that gets you in trouble is rarely invented. It's a real number with the wrong owner, like a client's result written up as yours. You sell programs, so a buyer reads a number in the post as a promise about what they'll get.

• Open the draft beside the video.
• Read the timestamp in brackets after each number, name and promise.
• Play 10 seconds of the video at that point.
• If the timestamp is missing or those 10 seconds don't contain the line, search the caption file for a word from it and write in the right time.
• Listen to each number as well. Captions mishear numbers, and your names list only covers names.
• Check who each fact belongs to, and that each name matches your names list.
• Resolve each [CHECK]: fix the line using the video, or delete it.
• Test: you can name a timestamp for each number in the post, and no [CHECK] is left.

Plan on about half an hour. That's our estimate for a 15-minute video with 8 facts to verify, at roughly 4 minutes each to find the timestamp and fix the line: 32 minutes. After your first three videos, use your own figure.

Send weak drafts back. If a draft needs more than light fixes, add any names the draft got wrong to your names list, add the section the draft wandered into (start and end time) to your skip ranges, then rerun Step 4. Rerunning the prompt costs less than rewriting the post by hand.

Step 6: Add headings, links and your offer

• Use your Step 1 question as the post title.
• Reword each heading as a question a reader would type.
• Check that the answer to your question sits in the first two sentences, right under the title.
• Paste the video embed below that answer.
• Add one link to your offer page in the last paragraph, with the program name as the link text.
• Test: read the page without playing the video. If it makes sense on its own, move on.

Step 7: Publish and log it

• Delete every timestamp note and confirm no [CHECK] is left.
• Confirm the offer link works.
• Click publish, then add the live URL and the date to the video's row in your log.
• Test: open the URL in a private window and confirm the page loads and the video plays.

When should you do it by hand vs automate it?

Do your first two or three videos by hand. You'll see where the draft goes wrong in your voice and on your topics.

Keep the Step 5 check with a person. Accuracy is one reason some small firms avoid AI, per a Federal Reserve Bank of San Francisco brief, and it's the reason this step stays with your reviewer. Don't skip it on any post, and on posts with client results, health, money or legal claims, have a second person read it too. Put the review in your calendar with a time and a reviewer's name. If the review lives only in your head, it's the step you'll skip in a busy week.

You can hand the rest to automation: the caption export, applying your names list, the first draft, formatting and the log row. Marking the tangents stays with a person, because it means watching the video. We'd only move a step over once it has gone the same way three times, because until then you don't know what to tell the automation to do. Our post on what content repurposing automation costs a coach shows what the finished version looks like.

What usually breaks?

Auto-captions may not be ready yet on a fresh upload. If YouTube Studio > Subtitles shows no automatic track for the video, come back later before you export. Don't type the transcript yourself.

The AI can turn a guest's or client's words into first person, so their story reads as yours. When you check who each fact belongs to in Step 5, put the speaker's name back on the line.

A misheard number still matches a timestamp. You said 15, the caption says 50, so the draft says 50 and the search finds it. That's why you listen to each number: play the line and take the number from the audio.

When your AI tool or blog platform changes, rerun the routine on a video you've already published and compare the two posts. If the prompt starts adding claims or the formatting breaks, fix it before your next upload. Our guide to an AI content workflow with a human approval step covers how to keep a person in the loop.

Where should you start with your first video?

Take your last published video and run Steps 1 to 5 by hand. It's one afternoon of work. By the end, you'll know how much cleanup your captions need and how long the check takes you.

If you'd rather hand this off, see what The On It Studio does or start with the free ops audit.