6 min read

522 Pages, 13 Clicks: Why Thin Pages Don’t Rank in SEO

Thirteen clicks. That was the total recorded in Search Console for one of our panels over 16 months, from early 2025 through September 2026. The panel had 522 indexed pages. It was not a new site without authority. It had clean URLs, correct technical structure, and published content. The problem was somewhere else entirely.

The Measured Symptom

When we looked at the full Search Console data for the period, the picture was clear: low impressions, near-zero CTR, positions oscillating between 30 and 60 for terms with real search volume. The 522 pages existed for Google, appeared in the index, but generated no traffic.

The number thirteen is not a typo. Thirteen organic clicks over 16 months for a set of more than five hundred URLs.

For anyone working in SEO, this pattern has a name. It is not a lack of external links. It is not a manual penalty. It is the expected behavior when thin pages accumulate at scale.

After measuring the similarity between the pages in this panel, we arrived at 92.2% average similarity. Each URL had a different structure, but the text was nearly identical: the same pattern repeated with variations of one or two words per page. Google did not see 522 distinct pieces of content. It saw one page replicated 522 times.

The direct consequence: all those pages were moved to noindex. Google signaled there was not enough value to keep them in the active index. Thirteen clicks in 16 months, and then technical silence.

The Conceptual Root Cause

A thin page is not just a short page. It is a page that does not solve search intent in a differentiated way. Google evaluates each URL by asking: “if I show this page to someone who searched for this term, does it deliver what the search promised?”

When 92.2% of the text is the same across pages on the same domain, the algorithm’s answer is clear: these pages are not distinct enough to receive different rankings. Instead of distributing traffic across 522 pages, Google picks the most relevant one and ignores the rest, or pushes them into positions with no real visibility.

There are three paths that lead to this state:

Volume generation without real differentiation. When content is created at scale with a fixed pattern and only superficial variables (like names or prices), each page ends up being a functional copy of the others. A high volume of URLs is misleading: it does not multiply ranking potential, it dilutes it.

Missing search intent. Even a page with unique text fails if the term it tries to capture does not correspond to a real question someone actually types. Pages built for invented or overly specific terms accumulate near-zero impressions because no one searches for those things.

Lack of engagement signals. Google uses real user behavior to confirm quality. A page that receives visits, has a low bounce rate, and a reasonable session time accumulates positive signals. A page that exists but is never visited accumulates nothing, and the algorithm loses reason to keep it well-ranked.

What we measured was the combination of all three: replicated text with minimal variation, terms with no real search volume, and zero historical engagement. The result was predictable.

What We Did

After measuring the problem with Search Console and similarity analysis, we made the decision not to scale that model.

We did not publish more pages with the same pattern. We did not try to “fix” the existing 522 pages by adding extra paragraphs to mask the similarity. That cosmetic approach does not solve the real problem: the structure did not differentiate content in a way Google considered relevant for specific searches.

The license that this result gave us was to recalibrate the creation criterion. The question that now guides the process is not “how many pages can we create?”, but “does each URL we create answer a different search in a genuinely different way?”

This means measuring before scaling. If the content of two pages is fundamentally the same, publishing both is not having two SEO assets. It is creating noise.

We also learned that noindex at scale is not the end of a content strategy. It is the diagnosis that was missing. With 92.2% measured similarity, the decision to consolidate or deactivate duplicate pages now has a technical basis, not just intuition.

Search Console spent 16 months pointing at the problem. The data was there. Thirteen clicks for 522 pages leaves no room for doubt about what Google is communicating.

The Lesson

The most common mistake in programmatic SEO is not technical. It is conceptual. The assumption that “more pages equals more traffic” only works if each additional page corresponds to a distinct search intent, with content that satisfies it in a differentiated way.

Thirteen clicks across 522 pages in 16 months is not a technical execution failure. It is the algorithm’s response to content that does not deliver what it promises for real searches.

The 92.2% similarity number is the diagnosis. The noindex at scale is the verdict. And the way back runs through measuring before creating, not after.

If you are building any content strategy at volume, three questions are worth asking before publishing each batch:

Does each URL answer a different search?
Is the text on each page genuinely different from the others, or does only one variable change?
Is there real search volume for the term each page is trying to capture?

If the answer is “no” to any of the three, the batch will accumulate impressions and zero clicks, exactly like the 522 pages we measured.

Content that does not differentiate does not rank. This is not a technical limitation of SEO. It is the definition of what the algorithm was designed to do: show the user the most relevant result, not the one published in the highest volume.

The data point of 13 clicks in 16 months is clear enough. Google does not need volume. It needs relevance.

Enjoyed this analysis? Subscribe to the newsletter to receive the next case studies directly in your inbox.