Your best-performing post has sat at position three for two years. It still pulls thousands of visits a month and still looks great in every report. But the demo bookings it used to send stopped a while ago. Traffic is there. Results are not. What changed? If that sounds familiar, you're looking at a content refresh SEO problem.
Nothing about the post changed. The world around it did. The answer it gives has aged, the tools it names have new names, and the buying criteria your customers use have moved. Google keeps ranking it because the topic is still relevant, and relevance is only half the job. The other half is whether the page still earns the click and still moves someone toward a decision.
What changed under a page that still ranks
Start with the numbers that explain why old posts dominate search. Ahrefs' study of how long it takes to rank found that 72.9% of pages in Google's top 10 are more than three years old, and the average page at position one is five years old. Only 1.74% of newly published pages reach the top 10 within a year, which makes the post you already have the cheapest ranking asset you own.
The same logic explains why publishing new content is a worse bet than refreshing an old post. In Ahrefs' 2023 search traffic study, 96.55% of pages in its index get zero traffic from Google. Publishing something new is a lottery ticket. Refreshing a page that already ranks is a repair job on an asset you can see.
What breaks inside that asset is freshness. Google's documentation on ranking results says its systems weigh freshness more heavily for queries where the answer changes over time. A post about a tool, a price or a process written in 2024 is quietly answering a 2024 question. The intent is the same. The facts are not.
Search Engine Journal calls the evergreen myth out directly, noting that evergreen content doesn't maintain itself and decays slowly without regular updates.
There is a second audience you lose when a post ages. Ahrefs' study of AI assistant citations analysed 17 million citations and found that URLs cited by AI assistants are 25.7% fresher than URLs in the organic results. A stale post can keep ranking in Google while AI engines reach for a newer answer from someone else. Content Marketing Institute argues the same point for AI search, where outdated content lives forever and keeps influencing answers long after it stops ranking. A refresh is a search play and an AI visibility play at the same time.
How a content refresh SEO workflow finds the posts worth saving
You need two things, and you probably already have both of them. The first is Google Search Console, the free tool that shows which searches bring people to your site. The second is your analytics platform, whether that is GA4 or anything else that tracks form fills and bookings. That's the whole toolkit.
Judge the post on its trend
Open the Performance report in Search Console and look at each post's clicks and average position over the last 12 months. The pattern that matters is falling clicks with a stable position. Position holds, so the post is still relevant. Clicks fall because fewer people choose it over the results around it. That gap is where the refresh opportunity lives.
Query-level data is getting harder to read, so judge posts on the page-level trend. Ahrefs analysed 22 billion clicks across nearly 900,000 Search Console properties and found that 46.77% of clicks came from anonymised queries in April 2025. You can't always see which search brought the traffic anymore, but you can always see which pages are losing it.
Sort by conversions, not traffic
Traffic is a vanity read when the goal is pipeline. Search Engine Journal's content refresh guide makes the priority call explicit. When two posts perform similarly, the one whose conversion rate is falling month over month should take priority over the one with flat traffic.
A post that still ranks and still converts is doing its job. A post that ranks and stops converting leaks money quietly. Which post gets your refresh time first? Start with the one that ranks, still matches what you sell, and has started slipping.
Before you rewrite anything, read the post as if you were the reader today. Which stats are dead? Which tools have new names or new pricing? Which screenshots show an interface that no longer exists? Then compare the post against what a fresh search returns now. The gaps between the old post and today's answer are your edit list.
Update the answer, the proof and the call to action
Ahrefs runs refreshes on its own blog and publishes the method. When it rewrote and republished a 2018 post in 2024, traffic tripled, a 302% uplift (Ahrefs' republishing guide). The detail that matters is that the post was actually rewritten. Ahrefs is explicit that a date-only change does not count as a refresh. The method has three parts, and each one earns its place.
Update the answer
Rewrite the body around today's question. Keep the intent that earned the rankings, then update the facts so the tools, the prices and the process match today. Lead the first paragraphs with a direct answer, because that is the format Ahrefs' AI Overviews guide says AI engines quote. If you already run an AI writing agent, the drafting leg is the same workflow we described in our AI blog writing agent guide, applied to an existing post instead of a blank page.
Update the proof
Swap dead statistics for sourced ones. Name the studies, link the sources, date the claims. Freshness matters twice here. Ahrefs' AI Overview citation research found 38% of citations pull from the top 10, so a refreshed page that holds its ranking also holds its AI visibility. Screenshots and examples should be current as well, because a reader who spots a stale interface doubts everything else on the page.
Update the call to action
The offer your reader sees at the bottom of the post probably changed since you wrote it. Update the call to action so it matches what you sell today, and link it to the page that turns readers into conversations. Internal links matter here too. Point the refreshed post at your current resources so the pipeline it feeds still exists. On our site, the content repurposing workflow is where refreshed posts become social posts, so the two belong together.
Measuring whether a refresh worked
Give the refreshed post six to eight weeks, then compare the same numbers you used to pick it, clicks, position and conversions in analytics. Search Console keeps the history on the same URL, so the before and after sit side by side. Ahrefs tracks this internally with its Content Changes feature, and one documented overhaul added over 3,000 visits a month. You can track the same thing with a simple before-and-after note, or in a reporting dashboard if you already have one.
Search Console also added a signal that did not exist when most refresh advice was written. In June 2026 Google introduced Search Generative AI performance reports, which show whether your pages surface in AI answers. After a refresh, this report tells you whether the AI visibility half of the work landed.
Be honest about what a refresh can prove. There is no credible public benchmark for how much a content refresh lifts conversion rates. Most numbers in circulation come from vendors selling refresh services, so measure your own baseline before you touch anything and judge the workflow on your own trend. Ahrefs makes a related point about zero-click search. Not every ranking earns a click these days, and a refreshed page that earns mentions and influence is still doing work even when the click does not come.
When to leave a post alone
Some old posts should not be refreshed at all. Google's John Mueller has drawn the line between old content and bad content, and the two need different treatment (Search Engine Journal's coverage of his guidance). If the post is old but accurate and useful, refresh it. If it is thin or pointless, remove it or merge it into a stronger page.
Google's helpful content system rewards pages created primarily for people, and its spam policies are explicit that mass-produced content aimed at rankings, including AI-generated refresh spam, violates the rules. Google's AI content guidance says the same thing from the other direction, rewarding quality regardless of how the content was produced. Refresh the posts that deserve a reader, and leave the others alone.
What you can do this week
Which post gets your refresh time first? This is how you answer that in the next few days.
- Pull 12 months of clicks and position from Search Console for your blog posts, and mark the ones where clicks fell while position held.
- Pick the single post that ranks, still matches what you sell, and has started slipping. One post done properly beats five posts that each get a date bump.
- Run the three-part refresh, updating the answer, the proof and the call to action. Use an AI assistant for the drafting and make the judgement calls yourself.
- Build the business case while you work. If one new customer is worth AUD 2,000 to you, and a refreshed post brings two extra customers a month, that is AUD 4,000 a month and AUD 48,000 a year for a few focused hours. When the number beats the cost of your time, the workflow pays for itself.
This is something we do at Supernodes. Two-week pilot covering audit, refresh and measure. Speak with us if this sounds like your Monday morning.
Frequently asked questions
How often should I refresh old blog posts?
There is no fixed calendar. Refresh when the data says a post is losing momentum, clicks down, conversions down, or the answer it gives has aged. Some teams run a quarterly pass over their top pages, and others refresh on demand. What matters is that the refresh changes the substance. Google's guidance on helpful content frames the benchmark as content that benefits people.
Can I use AI to refresh my posts?
Yes, and Google's position is settled on this. Its AI content guidance says the quality of content matters more than how it is produced, as long as the work is genuinely helpful and does not try to manipulate rankings. The spam policies are equally clear that mass-producing refreshes for rankings is a violation. Use AI for the drafting labour, and keep the judgement human.
Will refreshing a post hurt its rankings?
A rewrite that changes the URL, the topic or the intent can cause a drop. A refresh that keeps the URL, keeps the core question and updates the answer, the proof and the call to action usually holds the rankings, because the page is still the best answer to the same question. A refresh is also the right moment to fix page experience issues such as mobile layout and slow loading.
Should I delete or merge posts instead of refreshing them?
Sometimes. If a post has no search demand and no business value, removing it or merging it into a stronger page is better than refreshing it. Search Engine Journal's guide to underperforming content walks through the refresh, consolidate or remove decision. Old content that is still accurate deserves a refresh, while thin pages that exist only for rankings are better removed.
Does a refresh help with AI search visibility?
It should. AI assistants retrieve current web data, and Ahrefs' LLM search analysis lists freshness among the factors that make brands visible to them. Its citation study found AI-cited URLs are 25.7% fresher than organic results, and Google's own AI Overviews lean on top-10 pages. A refreshed page that holds its ranking is therefore the same page most likely to be quoted.
How long does it take to run a refresh?
The foundation can be live in two weeks. The Supernodes pilot covers audit, refresh and measure, and most teams see the workflow running on their first one or two posts inside a fortnight. A single post refresh, done properly, is a few hours of focused work.