Content pruning is the least intuitive thing in search optimisation, because every instinct says more pages should mean more traffic. Publishing feels like accumulation. Deleting feels like throwing away work somebody paid for.

The mechanism that makes removal helpful is competition between your own pages. When two of yours target the same intent, they split the signals that would otherwise concentrate on one, and a retrieval system deciding between them may pick neither. The fix is not to improve both. It is to stop having two.

Before deleting anything, look at what the pages are doing individually. A page with no clicks may still hold a link that carries authority, or convert one visitor a month into an enquiry worth more than the rest of the site. Traffic is the wrong single criterion, and pruning done on traffic alone is how sites lose the pages that were quietly working.

What Content Pruning Is Actually Fixing

Three distinct problems, and they need different treatments.

Cannibalisation. Two or more pages competing for the same intent. This is the case where merging produces a genuine, measurable gain, because the combined page inherits the signals from both.

Thin or obsolete pages. Content that answers nothing anybody asks now: a post about a version nobody runs, an announcement from four years ago, a page generated to fill a content calendar. These rarely harm rankings directly, but they consume crawl attention and dilute what your site appears to be about.

Index bloat. Pages that should never have been indexed. Tag archives with one entry, paginated series, filtered views, staging URLs left crawlable. This is a technical problem masquerading as a content one, and the answer is usually a directive rather than a deletion.

Only the first is really about content quality. Diagnosing which of the three you have determines whether you merge, delete or configure, and doing the wrong one wastes the effort.

Deciding What Goes

Build one table with a row per page and these columns: impressions, clicks, the queries it ranks for, inbound links, conversions or enquiries attributed to it, and last meaningful update. Then work through four questions in order.

Does another page target the same intent? If yes, merge rather than delete. One page absorbs the other’s useful content, and the weaker URL redirects to it.

Does it have inbound links? If yes, do not delete it outright. Either keep it or redirect it somewhere genuinely relevant, because an irrelevant redirect is treated as roughly equivalent to a page that is gone.

Does it convert, even rarely? A page with forty visits a month and one enquiry is among the most valuable things you own. Traffic reporting will never show you this.

Is it still true? An outdated page can be worse than no page, particularly where a retrieval system might quote it. Correcting a factual claim is cheaper than either keeping it wrong or removing the page.

What is left after those four questions, pages with no traffic, no links, no conversions and nothing accurate left to say, is the set you can remove without thinking hard about it.

Merging Beats Deleting

When two pages compete, consolidation is almost always better than removal, because you keep the accumulated value of both.

The mechanics are unglamorous. Pick the stronger URL, usually the one with more links rather than more traffic. Move across anything the weaker page said better. Redirect the weaker URL to the stronger one. Update internal links to point at the survivor rather than through the redirect, because a chain of redirects is a small tax paid on every crawl.

The judgement call is what to keep. The merged page should answer one intent completely, not both intents partially. If you find yourself unable to combine them without the result covering two subjects, that is evidence they were not competing and should stay separate.

This is the same principle as writing one page per question, covered in why content ranks but never gets cited: two half-answers lose to one complete one.

Doing It Without Losing Anything

Redirect, do not delete, wherever a page had any value. A permanent redirect to a relevant page passes most of the accumulated signal. A redirect to the homepage from a specific article does not, and is generally handled as a soft error. Google’s guidance on removing content covers which method suits which case.

Keep a record of what you removed and when. When traffic moves in the following weeks, you need to be able to attribute it. Without a log, the analysis is guesswork.

Change in batches, then wait. Removing four hundred pages at once makes the result uninterpretable. Sizeable batches with a few weeks between them let you tell whether it is working.

Expect the numbers to fall first. Removing pages removes their impressions, so totals drop before consolidation shows up. The metric worth watching is clicks and conversions on the surviving pages, not sitewide impressions.

What This Does Not Fix

Pruning removes interference. It does not create authority.

A site with 166 posts and money pages ranking in the twenties and thirties does not have a content volume problem, and no amount of tidying moves those pages. When we ran a gap analysis against every post on this site, we found just seven uncovered commercial queries worth 162 impressions in total, measured over the 28 days to early August 2026. The coverage was already there. The pages that needed help needed links, not neighbours.

That is the honest limit. Pruning is worth doing when you have genuine cannibalisation or a large tail of pages that answer nothing, and it is cheap enough to be worth the afternoon. It is not a growth strategy, and a site that prunes instead of earning citations has simply made a tidier version of the same problem. Our generative engine optimisation guide covers the part that does move.

Mecanik runs audits of this kind as part of our SEO audit work, and the recommendation is frequently to prune less than the client expected.

Frequently Asked Questions

Does deleting old blog posts improve SEO? Sometimes, and only for specific reasons. Removal helps when two of your pages compete for the same intent and split the signals that would otherwise concentrate on one. It also helps where a large tail of pages answers nothing anybody asks. It does nothing for a site whose real problem is a shortage of inbound links.

Should I delete or merge competing pages? Merge. Consolidation keeps the accumulated value of both pages, where deletion discards it. Pick the stronger URL, usually the one with more inbound links rather than more traffic, move across anything the weaker page said better, redirect the weaker URL, then update internal links to point at the survivor directly.

How do I decide which pages to prune? Build a table with impressions, clicks, ranking queries, inbound links, conversions and last update per page. Then ask in order: does another page target the same intent, does it have inbound links, does it convert even rarely, and is it still true. What survives all four with nothing to offer is safe to remove.

What happens to traffic after content pruning? Totals usually fall first, because removing pages removes their impressions. The metric that matters is clicks and conversions on the surviving pages rather than sitewide impressions. Work in batches with a few weeks between them, and keep a log of what was removed so changes can be attributed.

Should I redirect deleted pages to the homepage? No. A redirect from a specific article to the homepage is generally treated as a soft error and passes little value. Redirect to a genuinely relevant page, or leave the URL to return a not-found response if no relevant destination exists.