Many people talk about removing pages with «thin» content. I’ve read this advice a million times online.
Delete bad pages WITH NO traffic—get a traffic boost.
Especially relevant in the era of the Google HELPFUL Update.
You’ve heard these thin content stories, and so have I. Plenty of experts, all confidently saying the exact same thing.
But is that actually true?
Are there any experiments with a sample size of over 100 websites?
No, there are none online. And I want to share with you a HUGE amount of work that was done and yielded no results whatsoever.
If anyone tells you that you need to delete THIN content or zero-traffic content, feel free to send them this article.
A glimpse of the work performed on removing thin content.
What Is the Thin Content Filter? Theory.
Thin content, Google Panda, or Google Fred—there are many names for this.
The bottom line is that Google considers the content low quality for some reason, and if a site has many such pages, all pages suffer, demoting the entire host in the end. There is no ROI from such a site—no matter how much content you publish, it’s useless. The most detailed public manual is available at https://alexeytrudov.com/google-fred
This filter applies to large websites. If you have 100–200 pages, I don’t think it applies to you.
It all comes down to the ratio of bad pages to good ones. If the ratio is high and you feed Google a lot of «bad» pages, a penalty hits.
At least in theory 🙂
Where Did So Many Sites Come From?
Such an experiment was only possible because, besides standard white-hat sites, I also build doorway sites for humans (DFH). Content on them ranks for years if generated properly. The content is useful and readable.
An example of one site over 3 years of existence:
The bounce rate is high and time on site is low, but as you can see, this content still ranks in Google—nearly 5 million visits 🙂
Duration of the Experiment
From July 1, 2021 to April 18, 2022. Once I saw the lack of traffic growth, the experiment was closed.
The table shows the number of pages indexed by Google. As you can see, this number decreased.
In total, there were 2 waves of content removal:
First wave—736k posts deleted
Second wave—1.8M posts deleted
After that, I tracked page indexing.
How to Clean Up Bad Pages?
Go to Yandex Metrika -> Landing Pages -> Source -> Google Search:
Export the pages:
Keep only the pages that received more than 10 visits from Google over the past year. Delete the rest.
The second removal option is checking unvisited pages via Google Cache. If the bot hasn’t visited in over 40 days, delete them.
The second option is possible via a-parser.com using this preset:
A-parser itself has a steep learning curve, and it took me several years to grasp it reasonably well.
Results
First group of websites:
Second group of websites:
As you can see, there was no real improvement in traffic, even though I deleted 2.5 million pages 🙂 That was the vast majority of pages across these sites—sometimes up to 80% of the total.
Conclusions
- If thin content is defined as any article that doesn’t get traffic, that definition is wrong. I wouldn’t delete articles over 500 words.
- Google did not respond with positive traffic growth after removing 2.5 million pages across a large sample of sites, even when thin pages accounted for up to 80% of total site pages.
Article Sponsor.
Links from my guest posts :))