How Britney Muller deindexed 75% of Moz.com's site and saw success - Credo

One of the most cited index-management stories in SEO is the Moz cleanup associated with Britney Muller, in which a large share of the site was removed from Google's index and organic performance improved rather than collapsed. It sounds counterintuitive: how does showing search engines fewer of your pages make you more visible? The answer is index quality over index size, and the mechanism is worth understanding before you copy the tactic.

What actually happened

The move was a deliberate deindexing of low value pages, the kind of thin, near-duplicate, or auto-generated URLs that large sites accumulate over years. Crucially, these pages were not deleted. They were kept live for users where useful and removed from search results with the noindex directive. The reported outcome was that organic visibility held up and improved as the bloated index shrank to a core of pages that genuinely deserved to rank. It is a case study, not a controlled experiment, so treat it as a strong illustration of a principle rather than a guaranteed formula.

Why removing pages can lift performance

Two forces are at work. First, quality is assessed partly at the site level, so a large mass of weak pages can drag on how the whole domain is perceived. Trimming that mass concentrates the quality signal. Second, crawl attention is finite. When search engines spend time repeatedly fetching thin URLs that will never rank, the pages you care about get crawled and refreshed less often. Removing dead weight redirects that attention toward pages that can earn traffic. Our guide on improving crawling and indexing with noindex and robots directives goes deeper on that budget dynamic.

The critical technical detail: noindex is not a robots.txt block

The single most common way teams break this tactic is by confusing the tools. The noindex directive lives in the page's robots meta tag or the X-Robots-Tag HTTP header, and a crawler can only obey it if it is allowed to fetch the page and read the directive. If you also block the same URL in robots.txt, the crawler never sees the noindex, so the page can linger in the index with a bare listing. The rule is simple: to deindex a page, keep it crawlable and let the noindex be read, then only consider a robots.txt block much later once the page has dropped out. Our noindex glossary entry and the full meta robots and X-Robots-Tag reference spell out the exact syntax.

How to apply the lesson without hurting yourself

Deindexing at scale is a judgement call, so audit before you act. Identify pages that are thin, duplicative, or have had no organic traffic and no links over a long window, and separate them from seasonal pages, conversion pages, and cluster-supporting pages that look quiet but do real work. Apply noindex to the genuine low value set, keep those URLs crawlable, and update internal links so you are not funnelling authority into pages you just told search engines to ignore. Then monitor. Watch indexed-page counts fall in Search Console and watch whether your retained pages hold or gain. If visibility drops, you removed something valuable, and because you used noindex rather than deletion you can reverse it quickly.

The takeaway

The Moz story is not an argument for deindexing pages at random. It is an argument for treating your index as a curated set of your best work. Fewer strong pages, crawled and refreshed often and free of thin-content drag, can outperform a large index padded with URLs that were never going to rank. Audit first, use noindex deliberately, keep the pages crawlable, and measure the result.

Reading the signals after you deindex

Once the noindex tags are live, the feedback comes mainly from Google Search Console. Watch the Pages report, where the deindexed URLs should migrate over days to weeks into the excluded bucket labelled as excluded by noindex tag, which confirms the directive was seen. Keep an eye on total indexed pages trending down toward your intended core, and on the Performance report for your retained pages, which is the number that actually matters. A healthy outcome looks like a shrinking index alongside flat or rising clicks on the pages you kept. A warning sign is clicks falling on retained pages, which suggests you deindexed something that was quietly feeding the site through internal links or long tail queries. Because noindex is reversible, the safe play is to deindex in batches and let each batch settle before the next, so a mistake is small, visible, and easy to undo rather than a single sweeping change you cannot cleanly diagnose.

Source: https://www.getcredo.com/britney-muller/

Frequently asked questions