Content Refresh SEO: How to Decide What to Update First
There is a moment most site owners recognise. Rankings have slipped, traffic feels flat, and the instinct is to publish something new. Sometimes that is exactly the right call. But a surprising number of the pages already sitting on your site, slightly stale and sitting at position 14, will outperform a brand new post if you give them the right attention. The difficulty is knowing which is which. The signals are there if you know where to look.
On this page
- Pages ranked 11 to 20 with decent impressions offer the highest refresh value.
- A page losing position while impressions stay flat is a strong refresh candidate.
- Low impressions, low clicks, and never ranking above position 30 suggests new content.
- Prioritise refresh work by proximity to the top ten first.
Last updated
Why the choice matters more than most people think
Publishing new content costs time. So does a proper refresh. Spend that time on the wrong one and you have a lot of effort pointing in the wrong direction. The decision should not be a gut call based on how old the post feels. It should come from what the data is actually showing you.
Search Console is the starting point. Filter by page over a rolling six-month window and look at three things together, impressions, clicks and average position. Each combination tells a different story, and each one points toward a different action.
What the numbers are telling you
A page that is losing impressions while its position stays roughly stable has a demand problem. The searches that used to land on it have shifted. Either the topic has moved on, or the way people phrase the question has changed. No amount of rewriting will fix a demand problem if the underlying intent has evaporated. That page probably needs a rethink, not a refresh.
A page losing position while impressions stay flat is a different situation entirely. Google still considers it relevant to the topic. Competing pages have simply moved ahead of it, most likely because they have been updated, deepened, or earned links since your version went up. That page is a strong refresh candidate. The domain authority argument has already been won to a degree. What it needs now is better content.
The third pattern is the one people overlook most often. A page sitting between positions 11 and 20 with decent impression volume but almost no clicks is not a lost cause. It is already in the index, already associated with its topic, and it is one good update away from moving into the top ten. Writing a new post on the same query and hoping it climbs faster than the existing one is almost always slower, because the existing page has history.
The specific signals that point toward a refresh
Beyond position data, a few patterns reliably indicate that a page needs work rather than replacement.
- The post answers the headline question but stops there. Search results for most topics now favour pages that address the follow-up questions too, not just the main one.
- Key examples, statistics or tools mentioned in the post no longer exist or have been superseded. Stale specifics erode trust quickly, particularly for readers who know the subject.
- The post covers a topic that has genuinely evolved since it was written. A piece about AI tools from three years ago, for example, will be missing entire categories of product that now dominate the search results page it is competing on.
- The internal structure buries the answer. If the page takes three paragraphs to get to what someone searching the query actually wants to know, competing pages that lead with the answer will win the click.
None of these require a full rewrite. A targeted refresh that adds depth, updates the specifics and restructures around the primary intent will do more than a new post on the same topic will do for twelve months.
When new content is the right answer
Not everything is worth refreshing. If a page has low impressions, low clicks and has never ranked above position 30 for a topic you genuinely care about, it has probably not built any meaningful signal around that query. A refresh there is patching something that was never really working.
New content makes sense when you are targeting a genuinely different angle on a topic, when search intent has shifted so far from what the original post covered that rewriting it would mean replacing almost everything, or when the existing post has structural problems deep enough that a clean start would be faster. A post written for a topic nobody searches for is not helped by being updated.
There is also a practical ceiling on how much refresh work any site should carry at once. The pages most worth updating are the ones already close to the first page. A page at position 40 is a much longer project than one at position 12, and the return on the work reflects that. Prioritise by proximity to the top ten first.
Refresh priority by current position, the closer to page one, the more a refresh earns
A practical way to decide
Pull a list of your ten lowest-performing pages by clicks over the last six months. For each one, check whether it has ever ranked in the top 20 for its primary query. If it has, and impressions are still present, a refresh is almost always the better move. If it has never broken through and impressions are negligible, new content on a different angle is more likely to move the needle.
One thing that is easy to miss, a post that covered a topic well but was published during a period when the site had less authority can do significantly better now, simply by being updated. The update signals recency. It often prompts a crawl. And the content itself, which was already reasonable, gets re-evaluated in the context of a stronger domain. That combination can move a page several positions without writing much that is genuinely new.
The work involved in a proper refresh is real. Reading the current content critically, finding what is thin, checking what competitors at positions one to five are covering that you are not, updating the structure so the answer comes earlier, those things take time and attention. But they are usually faster than starting from scratch, and on a page that already has some history, the results tend to come sooner too.