1. Start with an audit, not a rewrite

Don’t repurpose at random. Open your archive and sort it into three piles, and be ruthless about it.

Pile one: posts that still get occasional traffic or comments but mention tools, prices, or advice that has clearly moved on. These are your best candidates — readers are already finding them, and they’re finding them slightly wrong. Pile two: posts that were good but never found an audience, usually because the title was vague or the timing was off. These need a new angle, not just a polish. Pile three: posts that are wrong at their core, or about something you no longer stand behind. Leave those alone, or take them down. No amount of AI polish turns a bad premise into a good post.

I do this audit with a simple spreadsheet: post title, last updated date, what still holds up, what has changed, and one honest line about whether the post deserves a second life. It takes an afternoon for a small site, and it stops you from “refreshing” posts that should have been left in peace. The audit is the work. Everything after it is easier.

2. Update what dates fastest, and leave what still works

Most old posts aren’t entirely out of date — they’re out of date in about five places. Screenshots from an old interface. A tool that has changed its name or its pricing. A statistic from three years ago. An example that references a trend nobody remembers. A link to a page that no longer exists.

That’s your update list, and it’s where AI earns its keep. Paste the old post in and ask it to flag anything that looks time-sensitive: dates, version numbers, prices, named tools, and claims that depend on a moment in time. It will catch things your eyes glide over, because you’ve read the post too many times to see it fresh. Then — and this matters — verify every flag yourself. AI is good at spotting what might be dated. It’s not reliable at telling you what replaced it. Check the tool’s current site. Check the current price. Replace the dead link with a live one, or cut the sentence entirely.

What you should not do is rewrite the whole post because a few details changed. The voice, the structure, and the examples that still work are the reason the post was worth saving. Keep them. A repurposed post should still sound like the person who wrote it the first time, just better informed.

3. Give the old post one new shape, not five

Here’s where repurposing goes wrong: people turn one old post into a thread, a newsletter, a video script, a carousel, and a checklist, all in one afternoon, and every version is thinner than the original. Pick one new shape that fits the post.

If the old post is a long guide, the new shape might be a short, updated follow-up: what changed since I wrote this, and what I’d do differently now. If it’s a list of tips, the new shape might be a worked example — one tip, applied properly, start to finish. If it’s a post that never ranked, the new shape might be the same advice with a sharper title and a clearer first paragraph, aimed at the question people actually type into search.

AI can draft that new shape from the old post in minutes. Feed it the original, tell it the new format and the one thing that has changed, and let it produce a rough version. Then edit it the way you’d edit any draft: cut the generic lines, add one specific detail only you would know, and read it aloud before you publish. One strong second life beats five weak copies, and your readers can tell the difference.

4. Let AI draft the update, but keep the facts on your desk

The practical workflow I use now looks like this. First, I paste the old post and my audit notes into the tool and ask for a revised draft that keeps the structure and voice, updates only the flagged items, and marks every change it made. That last instruction matters — I want a visible list of changes, not a silently rewritten post. Then I go through that list one item at a time, checking each updated fact against a real source, and I delete anything I can’t verify.

Second, I ask it to suggest a new title and opening paragraph, because old titles are often the weakest part of an old post. I almost never take its suggestion whole, but it usually shows me what’s vague about mine. A title that said “Some Thoughts on Editing” becomes “The Editing Checklist I Actually Run on Every Draft” — same post, much clearer promise.

Third, I add the one thing AI can’t add: what’s happened since. A reader comment that changed my mind. A tool I stopped recommending and why. A result from trying my own advice for a year. That paragraph is the difference between a refreshed post and a re-skinned one, and it’s the reason the repurposed version is worth publishing at all.

5. Publish it like it’s new, without pretending it is

When the updated post is ready, treat its launch seriously. Give it a proper excerpt, check that its internal links point to your current posts rather than your oldest ones, and link to it from one or two related newer posts so readers can find it. If your site has a guides list or a sitemap page, make sure the updated post is on it — there’s no point rescuing a post and then leaving it out of the listings where readers browse.

What I don’t do is change the URL or pretend the post was written yesterday. The original date is part of its history, and readers trust a post more when it says, honestly, “first published in 2024, updated in 2026” than when it hides its age. An updated date note at the top, one line, does the job. It tells a new reader this post is maintained, which is exactly the signal an old archive post needs to send.

The honest truth

Repurposing isn’t a shortcut to more content. Done badly, it’s how a site fills up with slightly different versions of the same post, and readers notice that faster than any algorithm does. Done well, it’s the most respectful thing you can do with your own archive: you admit that a good idea deserves better than to go stale, you fix what’s actually broken, and you give it one clear new reason to exist. Your archive already did the hard part — it proved the idea was worth writing. The update just has to prove it’s still worth reading.

Frequently Asked Questions