A slipping post has one of four causes, and they need four different responses — so the first job is diagnosis, not rewriting. Refreshing the wrong page in the wrong way is how sites spend a quarter on content updates and end up flat.
Changing the publish date is not a refresh. It is the most common version of this work and the least effective — a page does not recover because its date says it is newer.
The four causes
Identify which one you have before touching the page. Each leaves a different signature in Search Console.
| Cause | Signature | Fix |
|---|---|---|
| Out-competed | Steady decline, impressions holding, position drifting | Make the page genuinely better than what passed it |
| Outdated | Sharper decline, often after a known change in the subject | Update the facts. This is the real refresh |
| Cannibalised | Two of your own pages trading positions for one query | Consolidate or differentiate — do not rewrite either |
| Query shifted | Impressions fall while position holds | Demand moved. The page may be fine and the topic smaller |
A fifth possibility worth ruling out first: a technical fault. A page that lost everything at once, rather than declining, is usually a noindex, a canonical change, a redirect or a robots rule — not a content problem. Check that before anything else, using why isn't my page indexed.
Diagnosing it in Search Console
Ten minutes, and it decides everything that follows.
Compare two periods — the last three months against the same three months a year earlier. Filter to the single page.
Then read three things:
- Impressions. Falling impressions with stable position means demand moved, not that you lost. Falling position with stable impressions means you were out-competed.
- Queries. Which specific queries lost ground? A page that lost its main query has a different problem from one that lost a long tail.
- Other pages ranking for the same query. If another of your pages now appears for the query this one used to own, you have cannibalisation, and rewriting either page makes it worse.
Also look at the SERP itself. If an AI Overview now sits above the results for that query, some of the decline is structural rather than about your page — the mechanism is in SEO when AI answers the question first.
The refresh process
Seven steps. Steps one and two are the ones people skip, and skipping them is why refreshes fail.
Step 1 — Rule out a technical cause
A sudden total loss is technical. Check indexing status, canonical, robots rules and redirects before assuming anything about content.
Step 2 — Diagnose which of the four causes applies
Search Console, two periods, one page. Do not proceed until you can name the cause.
Step 3 — Check for cannibalisation before rewriting
If two of your pages compete for the query, decide which should own it. Consolidate the weaker into the stronger and redirect it, or differentiate them so they target genuinely different intents. Rewriting both is the common mistake and it deepens the problem.
Step 4 — Update what is actually wrong
Not a full rewrite. Facts that changed, figures that dated, screenshots showing an old interface, recommendations that are no longer right, links that now 404.
This is where most of the value is, and it is a smaller job than a rewrite.
Step 5 — Fill the gaps the page does not cover
Look at what now ranks above you and identify what it answers that you do not. Add those sections only where they genuinely belong — padding a page to match a competitor's word count is the failure mode this step invites.
Step 6 — Fix the structure
Answer-first sections. Self-contained passages. A table where there is a comparison. This alone often recovers ground, because much older content buries its answer under an introduction nobody asked for.
Step 7 — Relink it
Add contextual links from newer pages on the same subject. An older page that fell out of your internal structure loses standing for that reason alone, and it is the most overlooked step. Method in internal linking.
Then update dateModified — because you genuinely modified it. That is what the field is for, and it is a consequence of the work rather than a substitute for it.
What not to do
Changing the date without changing the content. It is the thing people mean by "refresh" most often, and it does nothing. Worse, a dateModified that moves while the content sits still is a claim your page cannot support.
Rewriting from scratch. You lose whatever was working. Most declining pages have a core that still ranks and a periphery that dated.
Changing the URL. Almost never justified on a page with history. If the URL is genuinely wrong, redirect properly and accept a temporary dip.
Refreshing everything. Most sites have a handful of pages worth the effort and a long tail that should be left alone or consolidated.
Adding words to hit a competitor's length. Length is not the variable. Coverage is, and they are different.
Refreshing a page that should be deleted. Some content has no audience left. Consolidating it into a stronger page, or removing it, beats maintaining it — and this is the option nobody offers because it is not billable.
Which pages to refresh
Prioritise by opportunity, not by how much a page has fallen.
The best candidates share three traits:
- Meaningful impressions still. The page is visible; it is losing the click or the position
- A subject that still has demand. Check the query trend, not just the page
- A gap you can genuinely close. You can see what is beating it and you can do better
Deprioritise: pages on topics that no longer matter, pages that were never good, and pages competing with something better of your own.
A workable cadence: review quarterly, refresh a handful, and consolidate a handful. Continuous small maintenance beats an annual project, which is also true of most things on this site.
Frequently asked questions
How often should I update old blog posts?
Review quarterly and act on a handful, rather than refreshing on a schedule. Pages worth updating are those still receiving impressions on subjects that still have demand — a page nobody sees does not become valuable by being edited.
Does updating the publish date improve rankings?
No. Changing a date without changing content does nothing, and a dateModified that moves while the page sits still is a claim the page cannot support. Update the field because you genuinely modified the content, not to signal freshness.
Should I rewrite an old post completely?
Rarely. Most declining pages have a core that still works and a periphery that dated. A full rewrite discards whatever was ranking. Update the facts, fill genuine gaps, fix the structure, and relink it.
What is keyword cannibalisation?
Two of your own pages competing for the same query, so neither establishes itself. The signature is both pages appearing for the query in Search Console, trading positions. The fix is consolidating one into the other or differentiating their intent — not rewriting both.
Why did my traffic drop but my rankings stay the same?
Demand moved, or the SERP changed. Falling impressions with a stable position means fewer people are searching, or an AI Overview now answers the query above the results. Neither is fixed by editing the page.
Should I change the URL when I refresh a post?
Almost never on a page with history. If the URL is genuinely wrong, redirect it properly with a 301 and expect a temporary dip. In most cases the cost exceeds the benefit.
How do I know if a page is worth refreshing?
Three traits: it still gets meaningful impressions, the subject still has demand, and you can see a gap you could genuinely close. Missing any of the three usually means consolidating or leaving it alone is the better option.
Is it better to update an old post or write a new one?
Update, if the page still has impressions and the query is the same. Write new if the query has genuinely split into a different intent. Writing a second page on the same query creates cannibalisation, which is a worse problem than the one you started with.
What to do next
Pick one declining page and compare the last three months against the same period a year earlier in Search Console, filtered to that page. Look at impressions and position separately.
If impressions fell and position held, the topic shrank and the page may not be the problem. That single distinction saves more wasted work than any other check here.
Related guides
- On-page SEO: the complete checklist — what to fix once you know the cause
- Meta titles and descriptions: write them for clicks — when the loss is clicks rather than position
- Internal linking: how to build topic clusters — the most overlooked refresh step
- Why isn't my page indexed? A diagnostic — ruling out a technical cause
- SEO when AI answers the question first — when the SERP changed, not your page
The short version
- Rule out a technical causeA sudden total loss is technical rather than editorial. Check indexing status, canonical, robots rules and redirects before assuming anything about the content.
- Diagnose which of the four causes appliesSearch Console, two comparable periods, filtered to one page. Do not proceed until you can name the cause: out-competed, outdated, cannibalised, or query shifted.
- Check for cannibalisation before rewritingIf two of your pages compete for the query, consolidate the weaker into the stronger and redirect it, or differentiate their intent. Rewriting both deepens the problem.
- Update what is actually wrongFacts that changed, figures that dated, screenshots of old interfaces, recommendations no longer right, links that now 404. Not a full rewrite.
- Fill the gaps the page does not coverIdentify what now outranks you and what it answers that you do not, then add those sections only where they genuinely belong.
- Fix the structureAnswer-first sections, self-contained passages, a table where there is a comparison. Older content often buries its answer under an introduction nobody asked for.
- Relink itAdd contextual links from newer pages on the same subject, then update dateModified because you genuinely modified the page.
Free: The 60-Minute Email Authentication Fix
A no-fluff checklist to set up SPF, DKIM & DMARC correctly and pass Gmail & Yahoo's sender requirements.
Muhammad Basim has worked in digital marketing since 2013, focused on email deliverability and AI-assisted content production. He is the author of The Email Deliverability Playbook and The Email Copywriting Playbook.
Related Articles
Email Deliverability: Why Authenticated Emails Still Land in Spam
Email deliverability is whether your message reaches the inbox rather than the spam folder, and it is decided by three factors: technical infrastructure, list quality, and sending behaviour. Authentication is one item inside the first factor. Getting it right is necessary, and it is nowhere near sufficient. That gap explains the most common complaint in […]
Linkable Assets: What Actually Earns Links
A link is a citation, and a citation requires that someone writing about your subject needed you to make their point. That is the whole mechanism, and it is the reason most "linkable content" earns nothing. The test, before you build anything: Could a writer covering this topic finish their sentence without referencing you? If […]
Link Outreach Emails That Get Replies
Outreach fails for exactly two reasons, and only one of them gets written about. The first is that the email was not worth replying to. That is the reason every guide addresses, and the advice — personalise, be brief, offer value — is correct and insufficient. The second is that the email never arrived. Nobody […]