Most content teams have a graveyard. It's the 40-page archive of posts published two years ago that nobody reads, that rank for nothing, and that somehow never make it onto any priority list because they're not obviously broken. They just exist. And quietly, they're probably hurting you.
Google evaluates sites holistically. A site where 60% of pages are thin, outdated, or irrelevant sends a different signal than a site where every page earns its place. Content decay is real — not just traffic decay, but quality perception decay. The question isn't whether to deal with your archive. It's how.
The Three Buckets: Keep, Refresh, Delete
Not every old post needs the same treatment. Run every underperforming page through three questions: Does it still get any meaningful traffic or rank in the top 50 for anything worth caring about? Does it cover a topic that's genuinely relevant to what your site is about? And is the core information still accurate, or has the subject fundamentally changed?
If the answer to all three is yes — keep it, maybe update a few facts, add some internal links, and leave it alone. If it's relevant but thin, outdated, or no longer ranking — refresh it. If it's irrelevant to your current site direction, gets no traffic, and you can't remember why you wrote it — delete it. That's the framework. It's not complicated, but it requires you to actually look at each page.
What Makes a Refresh Worth Doing
A refresh is worth doing when the underlying query is still valid but your existing answer is no longer the best one. This happens in two ways: either the topic has evolved (new data, changed best practices, new tools available), or your original article was just undercooked — didn't go deep enough, didn't cite good sources, didn't cover the angles that search intent actually demands.
A refresh isn't just adding a paragraph at the top that says 'updated for 2025.' That fools nobody. A real refresh means re-reading the article critically, identifying what's wrong or missing, and actually rewriting the sections that need it. Sometimes that's 20% of the article. Sometimes it's 80%. If you're rewriting more than 80%, it might be faster to start fresh and redirect the old URL.
The biggest refresh opportunities are usually posts that rank between position 8 and 20 for a target query. They're indexed, they have some authority, but they're not getting meaningful clicks. That's often a content quality problem, not a link problem. Improving the depth and accuracy of those articles is one of the highest-ROI moves in SEO.
When to Delete Instead
Deletion is underrated and under-used. There's a psychological attachment to content that makes it hard to delete — you put effort into it, it exists, removing it feels like admitting failure. Get over that. Content that consistently signals low quality to Google is costing you more than its absence would.
Delete when: the topic is completely irrelevant to your current site direction, the page has received fewer than 50 organic clicks in the last 12 months with no signs of life, and there's no reasonable refresh path that would make it competitive. Seasonal content from past years that's no longer accurate. Promotional posts for products you no longer sell. Opinion pieces from a different era of the site that contradict your current positioning. These should go.
When you delete, redirect to the most relevant existing page. If there's no relevant page, a 404 is fine. Don't redirect to the homepage just to avoid a 404 — Google knows what you're doing, and it's not a quality signal either way.
The Consolidation Case: Two Thin Posts Into One Good One
Sometimes the right answer is neither refresh nor delete — it's merge. If you have two or three posts covering the same topic from slightly different angles, and none of them are performing well, consider combining them into a single comprehensive article. Redirect the old URLs to the new consolidated page.
Consolidation works well for topic clusters where the spoke articles are too thin to stand alone but collectively cover the ground well. Pull the best material from each, write the connective tissue, add current sources, and publish as a new article. The combined authority often produces better rankings than any of the individual pieces had on their own.
How Often to Run This Process
Once a year is the minimum. Twice a year is better if you publish frequently. The audit doesn't need to be exhaustive every time — rank your pages by traffic, look at the bottom 25%, and make decisions about those. The top-performing content usually takes care of itself. The graveyard is where the leverage is.
Use when you're doing a full refresh of a major article. Getting a source-backed new draft as a starting point, then editing against your existing content, is often faster than trying to rewrite the original line by line. You get fresh citations, current data, and a clean structure to work from.
A lean site of 40 strong pages consistently outperforms a bloated site of 200 mixed-quality ones. Google is judging the whole, not just the parts.