The short version: pruning a blog means auditing every post by traffic, backlinks and conversions first, then merging or redirecting the salvageable ones and deleting the rest with 410 status codes, not blanket 301s to your homepage. Done this way, most sites see a small dip for two to four weeks followed by a rise in overall organic traffic within eight to twelve weeks, because Google starts trusting the whole site more. The bit nobody tells you is that some traffic loss is not just acceptable, it is the point.
Why pruning even needs a plan
Wait, let me fix that heading.
Why pruning even needs a plan
Most people prune a blog the way I used to prune my rose bushes before I learned how: they hack off whatever looks dead and hope for the best. Then traffic drops and they panic and put it all back.
Here’s what’s happening under the surface. Google crawls your whole site as one entity. If a third of your posts are thin, outdated, or duplicate the same keyword three different ways, that drags down how Google rates the average quality of your domain. This is not a rumour, it is roughly what Google’s own Danny Sullivan has said publicly about how the helpful content system works: it looks at the site as a whole, not just page by page. So a pile of weak posts can quietly cap how well your good posts perform too.
That means pruning isn’t really about tidying up. It’s about protecting the posts that make you money.
My own pruning story, with the actual numbers
When I sat down to rebuild my blog in 2022, after a few rough years where content had been an afterthought, I had 340 published posts going back to around 2011. A chunk of them were 300-word news snippets about social media platform updates that had aged like milk. Facebook features that don’t exist any more. “Top 10 apps of 2013” style listicles. You know the type.
I pulled every single URL into a spreadsheet with four columns: 12-month organic sessions from Google Search Console, referring domains from Ahrefs, whether it had ever generated an email signup or enquiry, and how old it was. Then I sorted by traffic ascending.
228 posts had brought in fewer than 5 sessions in the previous year. Of those, 41 still had decent backlinks pointing at them from other sites. The rest had nothing propping them up at all.
I deleted 94 posts outright. I merged 63 into bigger, better posts on the same topic. I kept 41 as redirect targets because of their backlinks, pointing those old URLs to the closest relevant live page rather than the homepage. And I rewrote 15 that had decent traffic potential but were badly outdated.
Total: 340 down to 190 pieces of content on the site.
In the first three weeks, organic traffic dropped by 18%. I won’t pretend that didn’t make my stomach drop a bit, even knowing what I knew. By week nine it had recovered past the original baseline. By week fourteen, overall organic sessions were up 34% compared to before the prune, and average time on page across the site had gone from around 1:40 to 3:12, because the posts left standing were simply better and people stayed longer.
The bit that makes people uncomfortable
Here’s the truth most posts on this topic dance around: you will very likely lose some traffic, and that traffic was probably rubbish anyway.
A post pulling in 40 monthly visits from a search term with zero commercial intent, no email signups, no enquiries, nothing, is not a traffic asset. It’s a number that makes your Google Analytics dashboard feel busy. Losing it costs you nothing real. What people are scared of losing is the feeling of having a big blog, the ego of the archive. That’s not a business reason to keep a post.
I’ve had clients refuse to delete posts that hadn’t earned them a single lead in three years, purely because “it still gets traffic.” Traffic to what end? If nobody on that page ever clicks through, signs up, or buys, you’re not running a blog, you’re running a museum. Museums don’t need pruning advice.
Step by step: the audit
Before you touch a single post, build the same kind of spreadsheet I did. You need, at minimum:
- Organic sessions for the last 12 months (Google Search Console or Google Analytics)
- Referring domains and backlinks pointing at each URL (Ahrefs, SEMrush, or even the free version of Ubersuggest for smaller sites)
- Conversions attributed to the page, whether that’s an email signup, a contact form fill, or a product click
- Publish date and last updated date
- Whether another post on your site targets the same or a near-identical keyword
Once that’s built, sort by traffic first, then flag anything with under, say, 10 monthly sessions if you’re a smaller blog, or under 50 if you’ve got real scale. That’s your candidate list, not your final delete list.
Then sort candidates into four buckets
This is where most pruning guides get vague. Don’t just have “keep or delete.” You need four categories.
Delete outright means the post has no traffic, no backlinks, no conversions, and the topic is either dead (that platform feature doesn’t exist any more) or so thin there’s nothing worth saving. Return a 410 Gone status code, not a 301 redirect. This is the part almost nobody does correctly. A 301 to your homepage tells Google “this content moved,” which is a lie if it didn’t move anywhere, and Google’s own John Mueller has said 410s can be dropped from the index faster and more cleanly than a redirect to an unrelated page. Use a plugin like Redirection on WordPress, or set it at server level if you’re on something else.
Merge means two or three posts cover the same ground and none of them is strong enough alone. Pick the best-performing URL, fold the useful bits of the others into it, and 301 redirect the losers to that single winning page. When I merged 63 posts down into roughly 20 stronger ones, several of those combined posts ended up ranking for keywords none of the originals had ever touched, simply because the new version was thorough enough to deserve the ranking.
Redirect and retire covers posts with no real content value left but decent backlinks. You’re not saving the content, you’re saving the link equity. Redirect to the closest relevant live page, never the homepage as a default dumping ground, because a mismatched redirect can hurt more than a clean 410.
Rewrite and keep is for posts that get some traffic, sit on a topic that still matters, but are outdated, thin, or just badly written by an earlier, less experienced version of you (I have plenty of those from 2014). Update the date, expand it, fix the internal links, and republish under the same URL.
How many should you cut
There’s no universal percentage, whatever some listicle tells you. But as a rough steer from working across dozens of client blogs: sites with fewer than 100 posts rarely need to cut more than 10-15%. Sites that have been running five years or more, especially ones that used to publish daily “news” style content, often find 30-40% of the archive is dead weight. If you’re finding you want to cut more than half, slow down and check your data twice, because that usually means your tagging or date filters are wrong somewhere, not that half your blog is worthless.
Timeline: what to expect week by week
Set expectations before you start, both for yourself and for anyone you report to.
- Week 1 to 3: a dip, typically 10-20% of organic sessions, as Google reprocesses the redirects and removals
- Week 4 to 8: traffic stabilises, often just below where it was before
- Week 9 to 14: traffic climbs past the original baseline as the remaining, stronger posts start ranking better and getting crawled more often
- Month 4 onward: this is where it compounds, because Googlebot spends its crawl budget on fewer, better pages instead of wading through hundreds of thin ones
If you’re not seeing recovery by week 12, go back and check your redirects for errors, because a broken or looping redirect chain is the single most common reason pruning goes wrong.
Mistakes I’ve watched people make
Redirecting everything to the homepage. This is the laziest possible move and Google treats mass homepage redirects as a soft 404 signal, which can damage trust rather than protect it.
Deleting in one enormous batch. If you’re cutting more than 50 posts, do it in stages of maybe 20-30 at a time over a few weeks, so you can see the impact on rankings before you commit to the next round. I learned this one the hard way on an earlier, smaller prune years back where I deleted 40 posts in one afternoon and couldn’t tell which specific removals had caused a ranking wobble on an unrelated page.
Forgetting internal links. If 15 other posts link to the page you just deleted, you’ve now got 15 broken internal links, which is its own small SEO problem. Run a crawl with Screaming Frog before you delete anything, not after.
Confusing “old” with “bad.” Some of my highest-traffic posts today are from 2016. Age isn’t the criteria. Performance and accuracy are.
A quick word on AI-generated churn
Since 2023 I’ve seen a fresh wave of blogs stuffed with AI-written filler posts published purely to hit a quota, and now those same site owners are realising they need to prune half of what they just paid to have written. If you’re building content strategy from scratch in 2026, save yourself the future pruning job: publish fewer, deeper posts now rather than a high volume of average ones you’ll have to cut in eighteen months anyway.
Frequently asked questions
Will pruning my blog hurt my rankings straight away?
Usually yes, briefly. Expect a dip of roughly 10-20% in organic sessions for two to four weeks as Google reprocesses your redirects and removals, followed by recovery and typically a net gain within two to three months if the prune was done.
Should I use 301 redirects or just delete old posts?
Only redirect if there’s a relevant live page to send that traffic to, or if the old post has real backlinks worth preserving. If neither applies, use a 410 Gone status instead of forcing a redirect to an unrelated page or your homepage, since a mismatched redirect can do more harm than a clean deletion.
How do I decide which posts to cut first?
Build a spreadsheet with 12-month organic sessions, referring domains, and any conversions each post has generated. Anything with near-zero traffic, no backlinks, and no conversions is your first, safest batch to cut.
How many posts is it safe to delete at once?
For anything beyond about 50 posts, cut in batches of 20-30 every few weeks rather than all at once. That way if something goes wrong, you can trace it back to the batch that caused it instead of guessing.
Related reading: My Rebuild in Public Numbers, the Month Traffic Crossed Half a Million and How to Build an AI Content Calendar That Gets Used (Not Just Built).
For the bigger picture, see my full guide to content marketing.
