Deleting hundreds of thin pages can lift a site's organic visibility

The short version

  • We audited a crypto-research platform in India where only about half its submitted pages were indexed, and the fix we proposed was to delete most of one subfolder.
  • One block of auto-generated pages made up the overwhelming majority of the sitemap and accounted for almost all of the not-indexed URLs. The editorial sections indexed near-perfectly.
  • The move was not to improve the thin pages one by one. It was to cut them, redirect the few that earned it, and let the authority the site had already earned concentrate.
  • Google's own guidance backs this, and another site we pruned saw a real rise in clicks afterwards.
  • Some pages are assets. Some are liabilities. A single site-wide number hides which is which.

When a site will not index, the instinct is to add: more content, more pages, more fixes. On this one, the honest recommendation ran the other way.

Thin pages drag down the whole site

Google judges quality at the site level, not page by page. In its guidance on building high-quality sites, Google’s Amit Singhal was direct about it.

Low-quality content on some parts of a website can impact the whole site’s rankings. Removing or merging shallow pages could eventually help the rankings of your higher-quality content.

So thin pages do more than fail to rank. They pull down the pages you care about.

They waste crawl budget too. When most of your URLs carry little value, Google spends its crawl on pages it will never index while the content that deserves attention waits behind them.

Find the problem by segmenting, not averaging

Split your index coverage by subfolder. The site-wide average hides the problem; the breakdown exposes it in one pass.

On the platform we audited, the headline number looked bad on its own: only about half of submitted URLs indexed. The subfolder view told the real story.

Field diagnosis

One section drags the whole site down

Share of submitted pages that Google indexes, by site section

Dashboard ~100%
Blog ~100%
On-chain metrics ~100%
Crypto events ~87%
Coin pages (auto-generated) about half
One subfolder of auto-generated pages was the overwhelming majority of the whole sitemap, and it indexed at well under half. Those pages accounted for almost every not-indexed URL on the domain, while every editorial section indexed near-perfectly. Index coverage by subfolder. Values approximate.

One bloated subfolder was the whole problem, and the hand-written content was already winning.

Google had even split its own verdict. It crawled most of the dead pages and chose not to index them, and it found the rest without bothering to crawl them. Either way, it had voted.

It’s a brand problem, not just an SEO one

The thin pages cost more than rankings. This platform sells fundamental crypto research, yet hundreds of auto-generated coin pages made it read like a commodity price feed. A research brand that ships a shallow, templated page for every token undercuts its own promise.

Cutting the inventory back to the coins the team actually covers in depth does two things at once. It removes the pages Google rejects, and it makes the site look like what it claims to be. That alignment is the expertise signal Google rewards.

Cutting beats fixing

When one section is most of your URLs, improving it page by page is a year of work for a problem that subtraction solves in a quarter. You do not need to make hundreds of thin pages good. Keep the few that earn their place and remove the rest.

The precedent is real. Another site we worked on, in personal finance, cut its indexed inventory hard and its daily organic clicks rose while indexing climbed to near-complete.

Precedent

Fewer pages, more visibility

Indexed pages

~12,000 ~3,000

Cut on purpose

Indexing rate

about half near-complete

Rose afterwards

A personal-finance site we pruned. Daily organic clicks rose over the same period. The direction was not subtle: fewer URLs, more value per URL, stronger performance. Client figures rounded and approximate.

The pattern holds at scale too. In a public case study, SEO researcher Koray Tugberk Gubur documented a domain registrar that restructured and pruned, then won a core update.

At scale, publicly documented

The same move on a much larger site

Not-indexed pages before
1.28M
Core update won
Dec 2025
His line stuck with us: every URL removed helped Google focus on what mattered. Koray Tugberk Gubur, public analysis.

How to prune without losing equity

Careless subtraction is just deletion. Here is the sequence we use.

How to prune without losing equity

Six steps to prune safely

  1. 01

    Classify by index status

    Take every URL in the bloated subfolder. The pages Google crawled and refused to index are the first to cut. It has already told you what it thinks.

  2. 02

    Keep what earns its place

    Only pages with real search demand and real depth. For a coin page, that means a token people search for, with original analysis, not API-pulled price data.

  3. 03

    404 the dead, 301 the few

    Thin, never-indexed pages hold no links or rankings to save, so a clean 404 is the honest signal that they are gone. Use a 301 only for the few removed pages that earned links or traffic, pointing each to its closest relevant parent.

  4. 04

    Rebuild the sitemap

    List only the survivors, then resubmit it in Search Console.

  5. 05

    Concentrate internal links

    Point them at the retained pages instead of spreading them thin across hundreds of URLs.

  6. 06

    Strengthen the survivors

    Make each one clear Google's own test: does it offer original information or analysis you cannot get elsewhere?

Google Search Central guidance on site quality, applied.

Steps three and four are the ones that go wrong most often. Mapping redirects at this scale is the same discipline as a replatform, so if you are cutting hundreds of URLs it is worth working through the website migration checklist before anything goes live.

When not to delete

This approach targets one thing: thin, auto-generated pages that Google already rejects. Leave anything that earns traffic or holds real editorial content alone; those pages are already doing their job. The aim is to clear dead weight, not to thin out work that ranks.

Two guardrails keep it safe. Before you remove a page, confirm it has not earned links or traffic; the rare ones that have get a 301 instead of a 404. And review the cut list with the site owner before it goes live, because you can only reverse a deletion at a cost.

What to do this week

The instinct when indexing is bad is to add more. With Google’s quality bar still rising, the stronger move is often to remove. Work through this, in order:

  • Pull your index coverage and break it down by subfolder, not as one average.
  • Flag any subfolder that is a large share of your URLs but indexes poorly.
  • Cut the pages Google has already crawled and refused to index.
  • 404 the dead, never-indexed pages; 301 only the few that earned links or traffic.
  • Rebuild and resubmit your sitemap with only the pages that earn their place.
  • Strengthen the survivors so each clears Google’s originality bar.

If you want the wider technical picture around this, the SEO best practices checklist is the working list we run on client sites, and pruning is one part of the technical SEO work we do every day.

Sources

  1. More guidance on building high-quality sitesGoogle Search Central, 2011
  2. A domain registrar with 1.28 million not-indexed pages that won the December 2025 core updateKoray Tugberk Gubur, public analysis

Further reading