Content refresh: how to update old posts for more traffic
Every blog that has been running for more than a year contains articles that used to bring traffic and quietly stopped. The instinct is to write something new, but on most small business sites the fastest available gain is updating what is already there. An article that reached position eleven once has already proved it can rank, which is more than any blank page can say.
Why old posts lose traffic
Content decay is normal and it happens for four distinct reasons, which is worth knowing because the fix differs.
The first is that the information genuinely aged. Prices changed, a tool was discontinued, the interface you described in eight steps now takes three. Readers notice this faster than search engines do, and they leave.
The second is that competitors published something better. Search results are relative, so an article can lose positions without changing at all, simply because three more thorough pages appeared.
The third is that the search intent shifted. A query that once wanted a definition now wants a comparison, and a page answering the old question slides down regardless of quality.
The fourth is internal: you published newer articles on adjacent topics, and now two of your own pages compete for the same query. Neither ranks as well as the original did alone.
The practical consequence is that a refresh is not a cosmetic edit. You are diagnosing which of these four happened, then responding to that specific cause.
Finding which posts to refresh
Search Console is the only tool you need for this, and the whole exercise takes about half an hour.
Start with declining pages. Compare the last three months against the same period a year earlier, sort by the biggest drop in clicks, and you have your first list. These are pages that were working and stopped.
Then find the striking distance queries. Filter for positions between eight and twenty with meaningful impressions. These pages are close enough that a genuine improvement can move them onto the first screen, and that move is where almost all the traffic is.
Look for high impressions with a low click rate. That combination usually means the page ranks but the title and description do not persuade anyone to click, which is the cheapest fix in this entire article.
Check for cannibalisation while you are there. If two of your pages appear for the same query, decide which one should own it, then either merge them or narrow the weaker one to a different angle.
Prioritise by business value, not by traffic. An article about a service you actually sell is worth refreshing before a popular general piece that brings readers who will never buy anything.
What to actually change
A refresh that only swaps the year in the title is a wasted afternoon. Search engines compare the page to the competitors currently ranking, so the question is what those pages answer that yours does not.
- Correct everything that is factually out of date: prices, tool names, screenshots, legal requirements, steps that changed.
- Add the sections your page is missing. Read the top three results and note the questions they answer. Usually there are two or three genuine gaps.
- Rewrite the introduction. It was probably written for a different reader than the one who now arrives from search, and it is the part that decides whether they stay.
- Improve the title and meta description if the click rate is weak. Be specific and include the year only if the topic genuinely changes annually.
- Add internal links both ways: from the refreshed article to your relevant service page, and from newer articles to this one.
- Remove the padding. Most old articles contain three paragraphs of throat clearing before the answer, and cutting them helps both readers and the page.
- Add a short answer near the top for the main question. This is what gets quoted in AI summaries and featured snippets, and it costs one paragraph.
What not to touch
Do not change the address. If the article ranks, its URL carries the accumulated history and any external links. If you absolutely must change it, set a permanent 301 redirect and expect a temporary dip.
Do not delete the old content and replace it with something unrelated. That is a new article wearing an old page's clothes, and it usually loses the rankings the page had.
Do not update the published date unless you genuinely changed the substance. Sites that bump every date monthly to look fresh get no benefit from it, and readers who recognise the trick trust the rest less.
When you have done a real update, it is fair and useful to show it. Keep the original publication date and add an updated date beside it, which is honest and gives the reader the information they wanted from the date in the first place.
Do not refresh everything at once. Ten pages a quarter, chosen properly, tells you far more than fifty pages changed in a week, because you will have no idea which change did what.
And do not refresh pages that were never worth having. Some articles get no impressions because nobody searches for the topic. Updating them changes nothing, and merging or removing them is the better call.
Measuring whether it worked
Note the baseline before you start: clicks, impressions, average position and click rate for that page over the previous three months. Without this, the whole exercise becomes a matter of opinion.
Request indexing in Search Console after publishing. It does not guarantee anything, but it usually shortens the wait from weeks to days.
Then wait. Four to eight weeks is the realistic window for a refresh to show its effect, and checking daily during the first week tells you nothing except that positions fluctuate.
Judge by clicks and position for the target queries, not by traffic to the whole blog, which moves for a hundred unrelated reasons.
Expect a mixed outcome. Roughly speaking, a good refreshing habit produces clear gains on some pages, no change on others, and the occasional small decline. The portfolio average is what matters.
Make it a routine rather than a project. One afternoon per quarter, spent on the five worst declines and the five closest striking distance queries, keeps an entire small business blog healthy, and it is consistently cheaper than producing the equivalent traffic from new articles.
FAQ
How often should I refresh old blog posts?+
For a small business blog a quarterly pass is the right rhythm, covering roughly five to ten articles each time, chosen from the biggest declines and the queries sitting just below the first screen. Trying to review everything continuously turns into a job nobody does, while an annual sweep leaves declining pages losing traffic for months before anyone notices. There are two exceptions worth handling immediately rather than waiting for the quarterly pass: anything containing prices or legal information that has changed, and any article describing a tool or process that no longer works the way you described it, because those damage credibility as soon as a reader arrives.
Should I change the publication date when I update an article?+
Only when the substance genuinely changed, and even then the better pattern is to keep the original publication date visible and add a separate updated date next to it, because that is honest and gives the reader exactly the information they were looking for. Sites that bump dates automatically to appear fresh gain nothing from it, since search engines evaluate the content itself rather than the number printed on the page, and readers who spot an article dated last week describing an interface from three years ago trust the whole site less. The one case where the date matters a great deal is topics where currency is the point, such as prices, tax rules or annual comparisons, and there the correct answer is to actually update the content rather than to relabel it.
Is it better to refresh an old post or write a new one?+
If a page has ever ranked, refreshing it is almost always faster and cheaper, because it already carries indexing history, internal links and whatever external links it has earned, and none of that transfers to a blank page. Writing something new makes sense in two situations: when the topic is genuinely different rather than an updated version of the same question, and when the existing article was built around an intent that no longer matches what searchers want, in which case a rewrite so total that nothing survives is effectively a new article anyway. A useful test is whether you would keep the same URL: if yes, refresh it; if the address no longer describes the content you intend to publish, write the new piece and decide separately what to do with the old one.