Why refreshing old content beats publishing new content
Your best untapped traffic opportunity is almost certainly a page you published eighteen months ago and stopped thinking about.
Across the managed accounts we run, a refreshed page returns roughly twice the traffic per hour invested compared with a new one. The reason is unglamorous: the refreshed page already has links, age, internal signals and a position to defend. A new page starts from nothing.
Finding the right pages
Sort by traffic at risk, not by percentage decline. A page that fell 60% from 200 monthly sessions is a rounding error. A page that fell 12% from 40,000 is your quarter.
The formula we use is straightforward: current monthly sessions multiplied by the decline rate over the last two quarters, weighted by conversion value where you have it. Rank descending, take the top twenty, and stop looking at the rest until those are done.
Diagnosing why a page declined
Four causes, four different fixes. Getting this wrong wastes the effort entirely.
Intent drift
The SERP changed shape. A query that used to return guides now returns tools, or product pages, or a video carousel. Look at what ranks today. If the page type no longer matches, editing the copy will not save it — you need a different page, or you need to let it go.
Stale facts
Prices, statistics, product names and screenshots that are visibly out of date. Cheapest possible fix, and frequently the whole problem. Update the numbers, update the review date, resubmit.
Coverage thinned by comparison
Competitors published something more complete. Compare subtopic coverage against the current top five and fill only the genuine gaps. Do not pad — padding lowers answer density and hurts you on the AI surface.
Structural non-retrievability
The page still ranks but is never cited, because no paragraph stands alone. This is increasingly the most common finding on pages that hold their Google position while losing overall traffic.
How to refresh without losing what works
- Keep the URL. Always. If the URL must change, redirect properly and expect a temporary dip.
- Keep the core intent. Improving how a page says something is safe. Changing what it is about resets it.
- Rewrite passages, not pages. Targeted edits preserve the sections that are already earning.
- Update the internal links pointing in. Anchor text elsewhere on your site should reflect what the page now says.
- Resubmit deliberately. Sitemap ping, IndexNow, and a URL inspection request. Do not wait for a crawl that may be months away.
When new content is still the right call
Refreshing is not always the answer. Publish new when you have no coverage of a cluster at all, when the intent genuinely requires a different page type, when you are entering a topic for the first time, or when the existing page is so misaligned that a rewrite is a rewrite in name only.
A working cadence
The split that works for most teams is roughly 40% of content hours on refresh, 60% on new. Teams with more than about two hundred published pages usually find the refresh share should be higher — and almost always discover, when they finally run an audit, that a third of what they own is dead weight nobody had the authority to delete.