Content Refresh: How to Pick Which Old Posts to Update (And What to Change)

A content refresh is an update to a post that is already published and already indexed, made so the post answers its search query better than it does today. You keep the URL, change the parts that are out of date or missing, and then measure whether rankings and clicks moved. It is not a rewrite, and it is not changing the date at the top of the page.

Most advice on refreshing covers the editing. The harder part comes earlier: choosing which posts are worth the time. Refresh the wrong post and a careful afternoon of work changes nothing, because the post was never close enough to the first page for an edit to matter.

What Counts as a Content Refresh (And What Doesn't)

Three things get called a refresh, and only one of them is.

| Action | What changes | Is it a refresh? | | --- | --- | --- | | Updating facts, adding a missing section, rewriting the title | The content itself | Yes | | Changing the "last updated" date and nothing else | Only the date | No | | Deleting the post or merging it into another one | The URL goes away | No, that is pruning |

Google is direct about the second row. Its guidance on helpful content lists this as a warning sign: "Are you changing the date of pages to make them seem fresh when the content has not substantially changed?" The same list asks whether you are adding or removing a lot of content because you believe it makes the site seem fresh, and answers its own question with "No, it won't."

The third row is a separate decision. If a post has no search demand left, or competes with another post of yours for the same query, content pruning is the better tool. A refresh assumes the post still deserves its own URL.

Which Posts Are Worth Refreshing?

Sort candidates by where they rank now, not by how old they are. Age tells you nothing useful: a three-year-old post holding position 4 should be left alone, and a two-month-old post stuck at position 18 may be the best candidate you have.

Open the Performance report in Search Console, switch to the Pages tab, and read the average position for each post over the last 28 days. In the table, that number is the average position for that specific URL. Then put each post into one of three groups.

| Average position | What it means | What to do | | --- | --- | --- | | 1–10 with a low click-through rate | People see the result and pick something else | Rewrite the title and description only | | 11–30 | Google considers the page relevant but not the best answer | Fill content gaps and add internal links | | Below 30 | The page is several result pages deep | Leave it for now |

The middle group is where a refresh earns its time. A post at position 14 already has the relevance signal; it needs to become a more complete answer. This is the same logic as finding quick win keywords, applied to a whole page instead of a single query.

The first group needs a lighter touch. If a post already ranks on page one and gets impressions without clicks, the body is doing its job and the snippet is not. Changing the body risks the ranking you have. The fix lives in the title and description, which is covered in high impressions, no clicks.

The third group is the one people get wrong. A post showing hundreds of impressions at position 60 looks like an opportunity because the impression number is large. It is not a refresh candidate. A new title cannot win a click from a result page the searcher never opens, and the post would have to climb dozens of positions before anyone saw it. We use these same three bands to choose refresh candidates on this blog, and the below-30 group is skipped no matter how many impressions it shows.

One more filter before you start: confirm the post is losing ground for a content reason. If it dropped because of an indexing or canonical problem, editing the text will not help. The checks for telling those apart are in the guide to content decay, which is the diagnosis step that comes before a refresh.

What Should You Actually Change?

Work in this order. The early steps are cheap and fix the most common problems, so stop as soon as the post is a complete answer again.

1. Correct everything that has gone out of date. Prices, version numbers, interface names, screenshots, and any sentence containing "currently" or a year. These are the easiest details to verify and the quickest to fix.
2. Search the query and read what ranks above you. List the sub-questions those pages answer that yours skips. Add those sections. Do not lengthen the sections you already have.
3. Move the answer to the top. If the post takes four paragraphs to reach the point, rewrite the opening so the direct answer comes first.
4. Check the format against the query. If the top results are step-by-step lists and yours is an essay, restructure it. If they include a comparison table and you have none, build one.
5. Add internal links in both directions. Link from the refreshed post to newer related posts, and link from two or three related posts back to it. An internal linking audit shows which of your posts have ended up with no links pointing at them.

Do not pad. Google's helpful content guidance says plainly that it has no preferred word count, so a refresh that adds 600 words of filler is a refresh that added nothing. The measure is whether a sub-question went from unanswered to answered.

Be careful with posts that share a topic. If the section you want to add is already the subject of another post on your site, link to that post instead of duplicating it. Two pages answering the same question compete with each other, which is the pattern described in keyword cannibalization.

What Should Stay the Same

The URL. Every internal link and backlink the post has earned points at that address. Changing the slug forces a redirect and puts those signals at risk for no gain.

The original publish date, if your site shows one. Google's documentation on dates says a page can provide a publication date, a last updated date, or both. Showing both is the honest option after a real update. It also says the date visible on the page should match the date in your structured data, so update dateModified when you update the visible one.

The sitemap's accuracy. Google states that it uses the lastmod value in a sitemap only if it is consistently and verifiably accurate, and that the value should reflect the last significant update to the page. It gives changing a copyright date as an example of an update that is not significant. Google also ignores the priority and changefreq values entirely, so there is nothing to tune there.

Parts of the post that already work. If one section is the reason the post ranks for its main query, leave that section alone and build around it.

Does Fresh Content Rank Better?

Sometimes, and only for certain queries. Google describes having "query deserves freshness" systems, designed to show fresher content for queries where it would be expected. Its own examples are a recently released movie and a search about an earthquake that just happened.

That is a statement about the query, not a general bonus for recent dates. A post explaining how a canonical tag works does not become more useful because it was edited last week. A post listing the current prices of five tools does. Before refreshing for freshness alone, ask whether a searcher typing that query expects this year's answer. If they do not, the refresh has to earn its ranking through completeness instead.

How to Tell Whether a Refresh Worked

Decide how you will measure before you publish the update, or you will end up reading the result you hoped for.

• Write down the before numbers. For the 28 days before the change, record the post's impressions, clicks, and average position. Note the date of the edit and exactly what you changed.
• Ask for a recrawl once. Use the URL Inspection tool to request indexing for a single post, or resubmit your sitemap if you updated several. Google says crawling can take anywhere from a few days to a few weeks, and that requesting a recrawl of the same URL multiple times will not make it happen faster.
• Compare equal periods. After 28 days, use the compare option in the Performance report to set the new period against the old one, filtered to that URL. Search Console allows one comparison at a time, and Google notes that filtering by query can change totals because anonymized queries are left out, so compare at the page level.
• Leave the post alone during the window. If you edit it again after ten days, you can no longer tell which change did what.

A refresh that moves nothing is still information. It usually means the gap was not in the content, and the next thing to examine is links or search intent. And if the position dropped while clicks held steady, read whether a dropping average position is actually bad before undoing the work, because a page that starts appearing for more queries can show a worse average while doing better overall.

Keeping this log in a spreadsheet works for a handful of posts. Traffic Monitor connects to Search Console with read-only access and shows clicks, impressions, CTR, and per-post performance updated daily, so the before-and-after comparison for a refreshed post does not depend on remembering to export the numbers on the right day. You can sign in with Google to see where your existing posts sit today.

The Short Version

Pick refresh candidates by position, not by age: posts ranking 11–30 get new content and links, posts on page one with weak click-through get a new title, and posts below 30 wait. Fix stale facts first, add only the sections that are missing, and keep the URL. Update the date only when the content really changed. Then record the numbers, wait 28 days, and compare.