I cover the old HCU(X) classifier, site-level quality algorithms, the counterbalancing of systems, and how the transition has been as smooth as a bed of razor blades…
This is the first post in a series about the March 2024 broad core update. The update was one of the largest and most complex core updates Google has ever launched, it took 45 days to roll out, multiple core systems were updated that reinforce each other, and more. Many sites saw a ton of volatility and movement during the update, with some sites reversing course while the update was rolling out. There is no way I can cover everything about this update in a single post, so I’m opting to write multiple posts instead.
This post will cover the transition of Google’s helpful content system into the core ranking system. There is a ton of confusion about how that works, how it impacts sites overall, how it impacts sites previously hit by the September HCU(X), and more. I hope this can clear some things up for site owners and SEOs while also letting Google know more about what I’m seeing across many sites I’m tracking that have been heavily impacted by the September HCU(X). It’s not pretty.
The old HCU classifier, the transition of the helpful content system to core, and the counter-balancing of systems:
I’ve been getting a lot of questions about the old HCU classifier, page-level impact versus site-level, how sites will be impacted, etc. This is really important to understand and I’m not sure Google has done a great job with explaining how it works. I’ll start with how it is supposed to work and then cover what I’m seeing based on tracking many sites over time (especially since the September helpful content update rolled out).
With the transition to core, the old HCU classifier as we knew it is supposed to be gone. That old HCU classifier was applied when Google’s helpful content system determined that a site had a significant amount of “unhelpful content” based on a site-level evaluation (so Google understanding all of the content on a site and how much was “unhelpful”).
If the negative classifier was applied, a site would often plummet in rankings. The drops were not small with the September HCU(X)… I have shared many screenshots over time and here are just two from GSC. They are scary for sure.
Well, with the March 2024 core update, the helpful content system was baked into Google’s core ranking system and the old classifier was supposed to be retired. Instead, multiple systems in Google’s core ranking systems would now assess the helpfulness of content. This is where many got confused. Since the helpful content system is part of core now, it’s not some type of separate hit to rankings. Instead, it’s part of the overall calculation between systems that are part of Google’s core ranking system.
You can think of it this way. Each system that’s part of Google’s core ranking system (represented by a small factory below) might come up with a score based on the job it’s tasked to evaluate/assess. Each system could support or reinforce a certain score from another factory OR the systems could be at odds and counterbalance each other. So, imagine a site that might have some positive things and receive positive scores from certain core systems. But maybe the systems whose job it is to assess the helpfulness of content end up scoring the site poorly. In that situation, the various systems would be counterbalancing each other and the final scoring might not be optimal for that site.
My core point, pun intended, is that site owners cannot think of the helpful content system as some type of separate system. And the various systems part of Google’s core ranking system can be in agreement, or not. And if they aren’t, one system can counterbalance other systems. In other words, the systems are not working in isolation. They are part of a bigger system that leverages the scores from all the systems in Google’s core ranking system.
The importance of site-level quality algorithms:
I have covered many times over the years the power of Google’s site-level quality algorithms. They are a big reason why a site can surge or tank like crazy during a broad core update. It’s not like every url on a site suddenly became high or low quality… It’s the site-level adjustment during a broad core update that’s impacting rankings across a site.
And now with the helpful content system baked into core, the “helpfulness of content” now contributes to those site-level quality algorithms. And based on how those systems and assessments worked with rollout of the March 2024 core update, they seem to be influencing Google’s site-level quality algorithms heavily. And I mean really heavily.
For example, most sites I’m tracking that were heavily impacted by the September HCU(X) saw big drops with the March core update. And that’s after being down 80%+ based on the September helpful content update. I shared a screenshot the other day of a site that’s down a whopping 97% since September 18, 2023. It’s crazy to see. And this site does not deserve that in my opinion. I might write another post about that soon.
Google has explained many times that it evaluates sites overall with broad core updates. And the latest information about this is in their recently-updated help document covering debugging search traffic drops. They say if you saw a large drop in position for a wide range of terms, you should evaluate your website overall (and NOT just individual pages). I have been saying this FOREVER in blog posts and presentations covering broad core updates. Don’t miss the forest for the trees. Improve your entire site over time.
How the HCU transition was supposed to go versus how it went:
Every morning during the March core update I ran the visibility numbers for 380 sites impacted heavily by the September HCU(X). I was hoping that I would see movement across those sites as Google’s core systems took over assessing the helpfulness of content. And no, I did not believe all would recover… Some definitely shouldn’t recover. But I did think that I would see some recovery for certain sites that have been working to improve, and maybe some would drop more if they didn’t. I did not expect to see what actually happened…
Every morning I saw nothing but more drops across those sites. It was insane to analyze, and I feel terrible for site owners impacted heavily by the September HCU(X). I truly believed at some point during the update, the old classifier would be dropped and Google’s systems for assessing the helpfulness of content would take over. Nope. Every morning showed more of a drop for those sites and I shared that each day via my “Good Morning Google Land!” threads on Twitter. I’m not saying the old classifier is definitely still being applied, but not one site showed any type of recovery at all.
You can see an example of the continued drop in my tweet below. Again, many of those sites were already destroyed by the September HCU(X) and now were dropping even more.
Wrapping Up and why you should submit feedback:
I hope this post helped you understand more about the transition of Google’s helpful content system to its core ranking system, how site-level quality algorithms are extremely important (especially with broad core updates), and how the transition did not yield the movement that many thought it would.
Google is now actively collecting feedback based on the March 2024 core update. Site owners can supply sample queries and landing page combinations based on sites that wrongly dropped (or continued to drop) with the update. And I know that many sites impacted by the September HCU(X) are submitting their feedback as well. If you feel you have been impacted heavily by the September HCU(X) and you have worked to improve but you aren’t seeing any movement in the right direction, then you should definitely submit feedback.
Google can refine its systems to address problems and it has done this in the past based on feedback. For example, I wrote a post in 2018 when ecommerce category pages were getting sucked into video carousels by mistake. One client lost between $600K and $800K over a 6-8 week period while that was happening. I wrote a case study about that, which got on Google’s radar. Danny Sullivan provided that case to the teams working on video carousels and the problem was indeed fixed about three to four weeks later. I’m not saying the HCU problems will definitely be fixed… but the more examples site owners and SEO can provide Google, the more Google’s teams can make the necessary adjustments.
Here is the part of my post when the fix was rolled out by Google. You can see clicks return when the category pages were in the 10 blue links versus the video carousel. It was great to see that roll out…
Regarding the March core update form, I’m sure it’s getting hammered right now. And my hope is the teams involved at Google are reviewing as many examples as they can and then roll out changes to correct some of the HCU madness. Again, not all sites should recover. But some definitely should see some type of recovery… Time will tell if Google implements bigger changes to those systems assessing the helpfulness of content. I’ll be tracking this closely and will share if I see any significant changes. Stay tuned.
GG
