Decision guide
Refresh old blog posts or write new ones
Updating usually wins for a blog that already has links and impressions. Writing new wins when the topic is missing. Here is a test to settle each post in about a minute.

Refresh, merge or retire: side by side
| Signal in the old post | Best move | Why |
|---|---|---|
| Gets impressions in Search Console, few clicks | Refresh title and intro | Google already shows it |
| Has referring domains, thin text | Rewrite, keep the URL | The links stay attached |
| Two posts on one question | Merge, redirect the weaker | Ends self-competition |
| No impressions, no links, off topic | Retire or noindex | It dilutes the archive |
| Topic you never covered | Write new | Nothing to refresh |
| Dated facts or prices | Update numbers and date | Stale detail loses trust |
Pull the right data from Search Console first
Opinions about old posts are cheap. The Pages and Queries reports are free, and they show what each post actually earned. Start there before you decide anything about a post's future.
Filter for sixteen months
Set the date range to sixteen months. That covers a full year plus the season before it. A shorter window hides the seasonal dips that make a steady post look dead. Export the Pages table with impressions, clicks and average position. Compare the first and last eight weeks of that window as well. A post that dropped sharply in the second half has a different problem from one that declined slowly all year.
Read the last twenty-eight days separately
Look at the most recent four weeks too. Pull the queries for your top pages and note which ones still bring impressions. A post whose queries have shifted toward a neighbouring topic may need a new heading rather than a new URL.
Mark the posts with impressions but few clicks
These posts already appear for the question. The title and the first screen decide whether anyone clicks. Rewrite those two elements before you touch the body, since a clearer promise often does more than a longer article. If the impressions themselves are falling, the topic may have moved on. Then the post needs the rewrite described under the decision tree, not a new headline.
A decision tree you can run on every post
Open Search Console and pick the last sixteen months. Sort the Pages report by impressions, then walk down the list asking four questions in order.
First, does the post still answer what people search for? If no, rewrite it. Second, does the URL have links from other sites? If yes, never change the address. Third, does a better post of yours cover the same ground? If yes, merge. Fourth, is the topic still worth the blog's time? If no, retire it.
- Impressions but few clicks: rewrite the title and the first screen
- Clicks falling year on year: check whether the question changed
- Links but no traffic: update the content and keep the URL
- Neither links nor traffic: merge it into a stronger post
Imagine a gardening blog silent since 2023. Of 40 posts, 6 bring nearly all the impressions, 9 have at least one linking domain, and 12 duplicate each other. The plan: rewrite the 6, update the 9, merge the 12 into 4, and retire the rest. This is an example, not a result anyone has measured.
Five ways to refresh a post without starting over
Each move keeps the address, so whatever links and history the post carries stay attached.
Rewrite the opening
Answer the question in the first paragraph, then add the context. Many old posts open with a long story that delays the answer by a full screen.
Add a worked example
A short example with plain numbers gives a reader a reason to stay. Mark it as an example and keep it next to the claim it illustrates.
Replace dead outbound links
Check every external link. A reference that now returns an error makes the whole article feel abandoned, and a quick swap restores some trust.
Merge two thin posts
When two posts answer the same question, fold the weaker one into the stronger and redirect it. Readers get one complete page instead of two partial ones.
Tighten the headings
Headings that state what a section claims help readers and crawlers find the answer. Rename the vague ones so each heading says something specific.

Why updating often beats publishing
An existing URL already carries history: indexing, impressions, maybe links. A new post starts from zero and must earn each of those.
Cost matters too. Rewriting a post of 1,200 words takes a fraction of the time of researching and writing a new one, and the page keeps its address.
The exception is the missing topic. If readers search for something you never covered, no refresh will help, and a new post is the only answer. Check Search Console queries and your site search for those gaps.
A refreshed post still needs time to show a change. Judge it over several weeks rather than the first few days, and note the date you published the update so the comparison stays clean.
A refresh you can finish in one afternoon
One post per session keeps the work honest. Pick the post with the most impressions and the weakest opening.
- 1
Rewrite the first screen
Rewrite the title tag, the H1 and the first paragraph. Keep the URL exactly as it is, since the links attached to it depend on that address.
- 2
Check every fact and figure
Open each number and its source. Replace anything you cannot confirm today, and remove what you can no longer find.
- 3
Update the date honestly
Change the visible date only when the content changed. A new date on an untouched post misleads readers and gains you nothing. If you add an update note, say what changed.
- 4
Log the change
Note the date, the reason and the queries you targeted. Check the results in Search Console after several weeks, not several days.
What a real refresh changes
- A new first paragraph that answers the question straight away
- Facts, screenshots and prices checked against today
- Dead outbound links replaced
- Two or three links to your other posts, using natural wording
- A visible updated date that matches the changes
- A reason for a reader to stay, such as a table or a worked example

Deleting or merging old posts is a one-way door
Removing a post takes seconds. Undoing it is harder. Once a URL disappears, every link pointing at it starts to leak value, and the only way back is to restore the page or redirect it.
When a post has links, redirect it
A post that earned links should never simply vanish. Redirect it to the closest page you keep, and write one sentence in your notes that explains the match. Redirecting everything to the homepage is the lazy option, and it tells search engines very little about what replaced what.
When a post has no value, let it go quietly
Some posts never drew a visitor and never earned a link. Retiring them is reasonable, and removing them or marking them noindex can tidy the archive. Still check the history first, because one forgotten link is enough to change the answer. Keep a short list of what you removed and why. In a year, that list shows whether the cleanup helped or merely made the archive smaller.
Old drafts written by a tool need a human pass
Some blogs were filled in a hurry with drafts from an AI tool. Those drafts often read smoothly. They still carry claims nobody checked. A refresh is the moment to slow down and verify every line that makes a claim. Many owners only notice the problem after a traffic drop, when one wrong figure is quoted back to them.
Check every claim again
Look for sentences that sound authoritative but cite nothing. Replace each one with a checked fact, or cut it. A short accurate post beats a long confident one that cannot be defended when a reader asks where a number came from. A useful test is whether you could defend the sentence in front of its source. If you cannot find the source, the sentence is a candidate for deletion.
Rewrite the generic paragraphs in your own words
Generic filler is easy to spot once you read the post as a stranger would. Rewrite those paragraphs around something only you know, such as a result from your own project or a mistake you learned from. That is what gives a refreshed post its reason to exist. Read the paragraph aloud. If it could sit unchanged in a dozen other posts on the same subject, it still needs your own detail.
Where links fit into the plan
Refreshed posts are the best targets for any new links, because a placement to a page you just improved can only help it. Pointing money at a stale page is the common mistake.
Google folded its helpful content system into core ranking with the March 2024 core update, so a weak archive can weigh on the whole site. That is another reason to prune before you promote.
Once the six strongest posts are rewritten, a small set of placements aimed at them is a reasonable next step. If traffic dropped after a platform change, fix the redirects first.
Long silences also change what readers expect, which we cover in what happens after 90 quiet days.