Audit my site →

Blog

Content Pruning: When Deleting Pages Helps Rankings

Deleting weak pages can help a site rank, but the mechanism is subtler than the folklore, and the folklore gets people to delete the wrong things.

It sounds like alchemy: delete pages from your website and watch your rankings improve. Subtraction producing growth. And like most things that sound like alchemy, it’s real but widely misunderstood, and the misunderstanding causes damage — usually to exactly the pages that should have been left alone.

Start with why it works at all, when it works. A search engine forms an impression of your site as a whole, not just of individual pages. If your site has thirty genuinely good pages and three hundred thin ones — near-empty tag archives, auto-generated location pages that differ only by city name, blog posts from years ago that say nothing in four hundred words — then the thin pages aren’t neutral. They’re the majority of the evidence. A crawler sampling your site mostly encounters filler, and the impression that forms is of a site that’s mostly filler. The good pages sit inside that impression like good paragraphs in a bloated essay. Pruning works the way editing works: removing the weak material changes what the remaining material adds up to.

There’s also a plainer, human mechanism. Thin pages don’t just dilute the crawler’s sample; they catch real visitors. Someone searches, lands on your emptiest page, finds nothing, and leaves with a formed opinion of your whole site. Every page that can rank is making promises on your behalf. A site with fewer, better pages makes fewer promises and keeps more of them.

But notice what the mechanism requires: pages that are actively weak, in meaningful quantity. This is where the folklore goes wrong. People hear “deleting pages helps” and reach for the most available metric — traffic — and delete whatever gets few visits. This confuses two very different kinds of page. A thin page gets no traffic because it deserves none. But plenty of excellent pages get little traffic too: the essay on a narrow topic that twelve people a year need desperately, the documentation page that closes deals when a prospect finds it, the reference that other sites link to even if searchers rarely land there. Low traffic is a symptom shared by pages that should die and pages that are doing quiet, important work. Pruning by traffic alone is how sites amputate healthy tissue.

The better question for each candidate page is not “does it get traffic?” but “would anyone be worse off if this vanished?” And there’s a decision that comes before delete, which the folklore skips: consolidate. Often the problem isn’t that a page is worthless but that its worth is scattered. You wrote about the same topic four times over five years, each post half-decent, all four competing with each other in search and none strong enough to win. The right move is to merge them into one page that’s actually good and redirect the old URLs to it — a 301 redirect, the permanent kind, which tells search engines the page has moved and lets the new page inherit what the old ones had earned, including links from elsewhere. Deleting those four pages outright, by contrast, burns whatever they’d accumulated. Consolidation is pruning for pages that contain something worth keeping; deletion is for pages that contain nothing.

When you do delete, do it honestly. A removed page should return a 404 or, better, a 410 — the HTTP status that says “gone on purpose.” What you shouldn’t do is redirect every dead page to your homepage. It feels tidy, but a redirect is a claim that the content moved somewhere, and when hundreds of unrelated pages all claim to have moved to the homepage, search engines learn to treat those redirects as the 404s they really are. You’ve added indirection without preserving anything.

I’ll admit my bias here: I built a tool that grades pages, and grading pages one at a time teaches you that most sites’ problems are concentrated, not spread evenly. When I scan a site and find pages with no H1, no meta description, and three sentences of content, those pages are usually not failed attempts at quality. They’re pages nobody ever intended a human to read — byproducts of a CMS, stubs someone meant to finish, archives generated by default. The site’s owner has often forgotten they exist. That’s the real value of doing a pruning pass, and it’s larger than the ranking effect: it forces you to hold an inventory of what you’ve published and notice how much of it happened by accident.

So here’s the shape of the practice. Once a year or so, list everything your site serves — everything, not just what’s in your navigation. For each page, ask what it’s for. Pages with a purpose and an audience stay, whatever their traffic. Pages whose purpose is now better served elsewhere get merged and redirected. Pages with no purpose get removed, cleanly. What’s left is a site where everything present is there on purpose. Search engines reward that, but they reward it because it’s the machine-legible shadow of something readers were already rewarding: a site that respects their time by only publishing things worth reading.

Test your own website

Type your website address and press Test. In about ten seconds you see a grade for your homepage. Then pick what to order. The first option is free.

More articles

← All posts