Updating Old Blog Posts: When a Refresh Beats a New Article

Update an old blog post when it has lost more than 20% of its organic traffic year over year — the threshold Ahrefs uses to flag content decay in a quarterly audit — and when the page still targets a query you could realistically win today. Everything above that line is noise; everything below it is a page that is quietly dying and worth an hour of your time.

That's the short version. The longer version is that most founders refresh the wrong posts, at the wrong time, for the wrong reason — usually by changing a date and calling it an update. This guide covers the signals that justify a refresh, the cases where writing something new beats updating, what Google has actually published about dates, and a triage system you can run in about 30 minutes a quarter.

Photo: Johan Larsson · BY

When to update old blog posts: the four signals that actually matter

Content decay is the gradual decline of a page's organic traffic and rankings over time — not a sudden drop from a penalty, but a slow bleed over months as competitors publish better answers and the search results reshuffle around them. Ahrefs defines it as exactly that, and recommends flagging pages down more than 20% year over year during quarterly audits.

But traffic decline alone isn't enough to justify a rewrite. Four signals together tell you whether a post is worth saving.

The most valuable position is a post ranking 6–20 for a query you already half-deserve. It has demonstrated relevance, it has some link equity, and it's one or two quality steps from page one. A post sitting at position 60 usually has a different problem — the topic was too competitive to begin with, or the page never matched intent. That's not a refresh candidate; that's a lesson.

Intent drift is the one founders miss most often. If you wrote a 900-word definition post two years ago and the top 10 is now full of comparison tables and pricing breakdowns, no amount of polishing your definitions will move it. The query changed jobs. Either rewrite the post to do the new job, or let it go.

When writing a new article beats a refresh

Refreshing is cheaper, but it isn't always the right move. Write something new when:

• The post ranks for a query you no longer want. Traffic that doesn't convert isn't worth maintenance.
• The old post is a different article wearing the same title. If you'd have to replace 80% of the body, you're writing a new piece anyway — and you may want the old URL to survive for a different query.
• You want to target a genuinely different keyword. Don't cannibalize. Two posts fighting over the same query is a common reason both underperform. If you're not sure which queries are actually winnable for your domain, our guide to finding long-tail keywords a new site can rank for walks through how to read a SERP for weakness before you commit.
• The post never got traffic in the first place. Ahrefs' study of 14 billion pages found that 96.55% of pages get zero traffic from Google, and another 1.94% get between one and ten monthly visits. A page that never ranked has no signal to recover. Refreshing it is a coin flip; deleting or consolidating it is often the better call.

That last point deserves emphasis, because it cuts against the instinct to fix everything. A refresh works by amplifying a signal that already exists. If there's no signal, you're not refreshing — you're rewriting from zero, and you should treat it with the same seriousness as a new article.

Why freshness matters more in 2026 than it did in 2023

The economics of updating changed when AI assistants became a meaningful referral source, because they visibly prefer recent pages.

Ahrefs analyzed 16.975 million cited URLs across ChatGPT, Perplexity, Gemini, Copilot, Google AI Overviews, and organic Google results. The finding: URLs cited by AI assistants average 1,064 days old versus 1,432 days for organic search results — 25.7% "fresher."

The per-assistant split matters more than the average:

Read that table twice. ChatGPT's citations skew dramatically newer than classic search results, while Google's AI Overviews pull from essentially the same age distribution as the organic SERP. So "does freshness matter?" has a different answer depending on which surface sends you traffic. If your audience researches in ChatGPT, a three-year-old post is competing against a two-and-a-half-year-old median. If they're in Google, age is much less of a handicap.

The practical read: freshness is a tiebreaker, not a ranking factor you can farm. Which brings us to the thing Google has been explicit about for years.

What Google actually says about updating dates

Google's own guidance is unusually direct here, and it's worth quoting rather than paraphrasing.

From Google Search Central's guidance on page dates: "If an article has been substantially changed, it can make sense to give it a fresh date and time. However, don't artificially freshen a story without adding significant information or some other compelling reason for the freshening."

The same page recommends showing when a page was updated: "If you update a page significantly, also update the visible date (and time, if you display that). If desired, you can show two dates: when a page was originally published and when it was updated." It also advises pairing that with datePublished and dateModified structured data so algorithms can read it.

And Google's creating helpful, reliable, people-first content documentation lists this among the questions that signal search-engine-first content: "Are you changing the date of pages to make them seem fresh when the content has not substantially changed?"

Put plainly: updating the date without updating the content is on Google's published list of things that indicate you're writing for search engines rather than people. The date is a description of what you did, not a lever.

This is the same failure mode we wrote about in what AI watermarking means for slop detection — optimizing the signal instead of the substance works until the platform learns to read the substance directly, and then it stops working all at once.

A 30-minute quarterly refresh triage

You don't need a content audit tool. You need Search Console and a spreadsheet, once a quarter.

1. Pull the comparison. Search Console → Performance → Date → Compare → "Last 3 months vs. same period last year." Switch to the Pages tab. Sort by click difference, ascending. The worst decliners float to the top.

2. Cut the list. Ignore anything with fewer than ~20 clicks in the earlier period — there's no signal to recover. You should be left with somewhere between three and fifteen URLs.

3. Classify each one. For every URL, open the query report filtered to that page, then search the primary query yourself. Assign one of four labels:

• Refresh — still the right query, still ranking 6–20, content is dated
• Rewrite — right topic, wrong angle; the SERP moved
• Consolidate — overlaps another post; merge and 301 the weaker URL
• Leave — decline is seasonal, or the query no longer matters to you

4. Do two per quarter. Not ten. A real refresh takes a couple of hours, and two done properly beats ten date changes. If you're the kind of founder who starts a content plan and abandons it by week three — the pattern we've written about at length — two is the number that actually survives contact with a shipping week.

What to actually change in a refresh

A refresh that moves rankings usually changes the shape of the article, not just its facts. In rough priority order:

• Rewrite the opening to answer the query in the first two sentences. This is the single highest-leverage edit, because it's what both featured snippets and AI assistants extract.
• Re-match the structure to the current SERP. If the top results all have a comparison table and you don't, add one. If they answer six sub-questions in H2s and you answer three, add three.
• Replace every stale number and re-verify the ones you keep. Open the source page. If the number isn't there anymore, cut it rather than carrying it forward.
• Add what's genuinely new since publication. New tools, new pricing, a changed regulation, a platform update.
• Fix the internal links. Old posts point at nothing; new posts don't point back. This is free and takes five minutes.
• Add or expand the FAQ. Standalone question-and-answer blocks are what assistants lift verbatim.
• Then, and only then, update the visible date and dateModified.

If you're using AI to draft the new sections, the quality bar is the same as for a new article — Google's position is about whether content is helpful, not how it was produced, which we covered in what Google actually says about AI-generated content.

Refresh vs. new article: the cost math

The reason refreshing wins so often is arithmetic. A refresh reuses the research, the URL, the links, and the existing rankings.

For what a full new article actually costs when you outsource it, we broke down 2026 freelance rates in how much a freelance SEO article costs. The short version is that the refresh is almost always the cheaper unit of output — which is exactly why it should be the default move on any post that still has a pulse.

Full disclosure: we build Rankody, an AI SEO autopilot that researches, writes, and schedules articles for founders — and every piece waits for your approval before it goes live. But a refresh queue is something you can run yourself with Search Console and a Tuesday afternoon. Tooling helps with volume; it doesn't decide which two posts deserve the work this quarter. You do.

FAQ

How often should I update old blog posts? Audit quarterly, but refresh selectively — two to four posts per quarter is a realistic pace for a solo founder. There's no calendar rule that says a post needs updating at six or twelve months. The trigger is a measurable traffic or ranking decline plus a query still worth winning, not the age of the file.

Does changing the publish date improve rankings? No, and Google explicitly warns against it. Google Search Central says not to "artificially freshen a story without adding significant information," and lists changing dates on unchanged pages among the signals of search-engine-first content. Update the date only after you've substantially changed the content.

Should I keep the same URL when I update a post? Yes, in almost every case. The existing URL carries the backlinks, the indexing history, and the rankings you're trying to rescue — changing it throws that away. Only change the URL if the topic has shifted so far that the old slug is actively misleading, and then 301 redirect the old one.

How long does a content refresh take to show results? Usually days to a few weeks, because the page is already indexed and Google recrawls known URLs faster than it discovers new ones. If nothing moves after about six weeks, the problem is probably intent mismatch rather than freshness — the SERP wants a different kind of page than the one you published.

Should I delete old posts that get no traffic instead of updating them? Often, yes. Ahrefs found 96.55% of pages get zero Google traffic, and a page with no historical signal has nothing for a refresh to amplify. Consolidating thin posts into one strong page and redirecting the rest is usually a better use of an afternoon than rewriting six pages nobody reads.

---

Refresh the posts that are decaying from a position worth defending, rewrite the ones whose SERP moved on without them, and leave the date alone until the content has genuinely changed.

Rankody researches your market, writes sourced long-form articles, and publishes on schedule — every piece waits for your approval first. Three articles free, no card. Analyze my site