I found one of the best examples of a true HCU recovery I’ve ever seen.

This site had an 83% organic traffic drop in the two months following Google’s Sept 2023 “Helpful Content Update”, and remained down 90%+ for more than a year, until the June 2025 Core Update when it shot up to new traffic highs.

What makes this example so interesting is that it’s a traditional “niche site” - an affiliate blog targeting “best of” buyer’s guide keywords. This is the type of site that Google punished the most with the HCU.

Self-sponsored, highly relevant ad

I found this recovery using Site Stats Database.

With monthly traffic data going back to September 2023 (month of the HCU), you can easily find every site that had traffic drops and recoveries between any two months.

And that’s just the tip of the iceberg.

You can also find thousands of low-DR, high-traffic sites in a matter of seconds.

Or high-traffic sites launched less than a year ago.

Or all sites monetized with Mediavine, Raptive, Journey by Mediavine, and more. Or a list of sites using LinkWhisper. Or any other plugin, theme, host, or technology.

It’s like BuiltWith, except every site includes monthly organic traffic numbers back to Sept 2023, Ahrefs DR, domain registration date, number of live pages, total backlinks and keywords, and so much more.

Easily one of the most underrated and powerful website and SEO research tools on the planet.

Explore it for free, or buy a plan for access to the entire database in a single sortable, filterable table.

When I look at the site, I can immediately tell that the owners put a TON of work into making it maximally helpful.

The site is: WaterFilterGuru.com (view all metrics for free in Site Stats Database)

Before the HCU in September 2023, this site was a typical buyer’s guide “niche site”. Although a bit better than most, as the guy behind it, Brian Campbell, had his face on the homepage (just like today), and an accompanying YouTube channel reviewing water filters.

But that wasn’t enough to spare the site during the HCU carnage.

So, how did this site recover, and even hit new traffic highs?

To find out, I dove into the Wayback Machine to find every meaningful change made to the site since the HCU. Here’s every change I found (in no particular order):

Google publicly claims it doesn’t matter whether you use AI to create content, as long as the content is original, high-quality, and people-first. But in the same guidelines, there’s this statement:

Should I use AI to generate content?

If you see AI as an essential way to help you produce content that is helpful and original, it might be useful to consider. If you see AI as an inexpensive, easy way to game search engine rankings, then no.

In the case of product reviews, AI is obviously not essential for producing helpful, original content. Most, if not all, people using AI to produce reviews are doing so to quickly pump out content (“game” rankings). It can work, especially in search engines like Bing/DDG/Yahoo, but no one in their right mind would honestly claim AI-written product reviews are more helpful and original than human-written reviews of products the writer actually tested.

Besides that, no one wants to read AI content. Even if it’s “high quality”. So, that “written by humans, not AI robots” statement is very reassuring to read from the perspective of an actual visitor trying to research water filters. And this means user signals like dwell time, scroll depth, and page interactions will be higher, as the users are made to feel comfortable and confident reading content that’s not AI-generated. User signals are absolutely a measure of quality, and correlate strongly with ranking.

This is the most drastic change after the HCU drop, and my hunch is that this is the primary contributor to the subsequent recovery nearly two years later.

There were two major article overhauls made. The first one was around October 21st, when actual images of Brian using and testing the products were added, and the content was completely rewritten with less generic wording and more of a “we actually tested this” tone.

The second major overhaul was made around April 14th, 2024. This time, the articles were again completely rewritten, with much denser information: actual test results, the addition of tables and bar charts showing contaminant levels and other measured data from the tests, short self-hosted/native videos (not embedded from YouTube or any other third-party) added throughout the articles showing the actual testing process and products in use, and completely revamped product overview boxes.

At this point, the articles were genuinely helpful, people-first, and the author had obviously tested the actual products in the review.

Brian actually became a Certified Water Specialist and Certified Water Treatment Representative by the Water Quality Association in February 2024.

I personally verified these weren’t fake credentials (not that I suspected anything, but niche site people are sometimes known to make stuff like that up ugh). Here’s his listing in WQA’s certified professionals directory:

Once he was certified, he added the credentials to his author bio on the site.

EEAT is not technically a ranking factor, but it absolutely does indirectly affect rankings. It’s a major signal Search Quality Raters look for when evaluating content.

This was a very interesting find. I don’t know how much impact it had, or whether it had any impact at all, but in my opinion (as a site visitor/reader), the updated titles seem more honest and friendly (people-first, not search engine-first).

The titles actually went through several changes since the HCU.

They started out as the perfectly search engine optimized “best of” format that was extremely popular pre-HCU:

“7 Best Water Filter Pitchers (October 2023)”

(odd number, “best”, [product], [month/year] - and optionally “buyer’s guide” at the end - was the template just about every affiliate site used).

Then, the format changed to: “The 7 Best Water Filter Pitchers for Safer Drinking Water of 2024”.

Not long after, it changed again to: “The 5 Best Water Filter Pitchers, According to Our Testing”.

Still “optimized”, but longer, without the misleading auto-updating date, and with a new EEAT-smelling line (“according to our testing”).

And as of last week, the format was changed once again, this time to: “The Best Water Filter Pitchers We’ve Personally Tested”. This latest change had no impact on recovery since it’s been a year since the recovery, but it’s still interesting to see!

At some point between May 19th and July 18th 2024, Brian added a consulting service to help people find good water treatment solutions.

This turned the site into an actual service business, not just a content blog. During the HCU, many sites survived that had a reason for existing beyond providing informational content. Consulting businesses, ecommerce sites, and local businesses were largely unscathed by the update. So, adding consulting was a smart move. Is it what caused the recovery? Who knows, but it certainly didn’t hurt, and will likely make the site more resilient come future updates.

In early 2025, still months before the recovery, Water Filter Guru launched a Shopify store (on a subdomain) selling hats and t-shirts.

Like I said, many ecommerce sites were spared in the HCU. Like consulting, this was also a smart move, but impossible to say whether it was a main driver in the recovery.

Here’s a complete timeline of the changes made between the HCU drop and the June 2025 recovery. This timeline is centered around a single article, that I chose as representative of the rest of the articles on the site (which were also updated similarly), and it includes sitewide changes as well.

This timeline includes all the major changes I mentioned in the previous section, along with other minor changes. You can view the Diffchecker links to see the actual changes between each version of the article if you want to verify my findings:

It’s impossible to say which of these changes had the most impact (or whether some of them had no impact at all), as they were all implemented prior to the recovery. Whatever the case, we can know for sure that on the whole, these changes brought the site back from the dead. I can say that with some level of confidence because I’ve seen many examples of sites that made no improvements post-HCU, and they remain dead. I haven’t seen any examples of sites that made zero changes and still recovered.

This was a fun newsletter to research and write. The conclusion is that Google rewards true people-first content (and perhaps a real product/service, not just info content). Nothing revolutionary, sorry, but I hope it’s helpful and inspiring to see a real-world example of a people-first content site that recovered from the HCU.

Let me know what you thought of this email here so I know whether I should consider writing more of these in the future:

Thanks for reading! Feel free to email back with questions and comments.

Ian

P.S.

I found this recovery, and many others, in less than a minute using Site Stats Database. I set the filters to:

Over 1,000 websites match this filter combination. Not all are true HCU recoveries, but a good portion of them are.

If you want to explore HCU recoveries yourself, Site Stats Database is the only tool in the world that can give you this data quickly and in bulk.

You can view the available plans here.

Have fun exploring! 😎