Google Algorithm Updates - 2024 Ongoing Discussion

[bernard]

Damn. I just checked out one of their articles! https://www.rtings.com/soundbar/reviews/hisense/hs2100 HOLY SHIT! That's better than ConsumerReports content quality! They fucking tested it at a scientific methodology level!

That's great.

I'm personally not a fan of "scientific" testing and reasoning in everything, but what this does is create a ton of unique fresh data, which is what Google wants, not rehashes of its old data. And of course the comments.

If it was a new domain with new links. No way would this rank.

I have a fresh domain site that ranks incredible with new content, better than the expired domain it is modelled over, a spinoff site, but it has links from the expired domain.

My expired domain site has consistently been "punished" when I've tried to move outside its strict contextual niche. This is definitely a thing for me. Simply isn't allowed with this domain, but for its niche keywords, it still ranks very well.

 

[BakerStreet]

They can talk all day about creating value for users... and YES their content is fucking awesome... BUT, they skipped over their biggest asset IMO, which is the fact that people are coming back directly to this website and engaging in their forum, creating content (UGC), and showing Google this isn't a spam website that is just trying to rank affiliate links.

That's what branding is. I can go deeper on this if anyone wants. If there's 5 likes, I'll sit down and write about it. It is based upon theory and experience and what I'm experiencing now. Not just with users but also bankers and the public in general.

 

[Ryuzaki]

HCU Classifier is Now Page-Level Instead of Sitewide

Marie Haynes has broken down the updated language in Google's official doc's related to the Helpful Content Updates.

The key takeaway if you don't want to read it is that the HCU classifier is no longer a site-wide signal and will affect individual pages now. Which means there's a chance that people will begin to see some pages pop back up while others remain tanked.

This will allow for some actual analysis and comparison to be done finally, and possibly a lil of the ole reverse engineering.

 

[Smith]

@Ryuzaki, great share. Thank you.

I just read Marie's article. She's a beast and I always value her thinking on search. But, like all of us, she's just putting out best guesses... though she stays very informed compared to most people.

I've looked at her "brainstorming" sessions before to get an objective eye on my sites. @Ryuzaki do you know anyone that has used this service? If so (or not), do you have any thoughts on engaging her for this sort of brainstorming?

This part here...

the HCU classifier is no longer a site-wide signal

... will be VERY interesting to watch if this is true.

That said, while I would personally love to see this change and will gladly take any rebound Google throws my way, I can't see sites that have been nailed during the Sept '23 and Mar '24 HCU updates getting a big rebound from this change.

If anything, maybe this change (from site-wide to page-level HCU classification) clears the way for sites to recover after making site-wide updates to E-E-A-T signals, and then page-level changes to reflect search intent, first-hand experience, and to deliver immediate value to the user.

If I'm reading the tea leaves correctly, site owners (e.g. me personally) should be rushing to get these site-wide E-E-A-T and page-level updates in place before their "Site Reputation Abuse" policy comes into effect on May 5.

It's totally possible that the May 5 deadline is irrelevant to most sites... unless you're a spam site with hundreds of third parties posting shit content... but it feels like Google is about to drop the hammer and I'm anticipating a lot more than just low-value third-party content getting slapped.

 

[Ryuzaki]

do you know anyone that has used this service? If so (or not), do you have any thoughts on engaging her for this sort of brainstorming?

I don’t know anyone that has used her and I don’t have a positive opinion that she knows anything more than anyone else. I see her as a neutral who used social media to position herself as an EEAT expert when it was new. She was loud and present and confident.

I always try to distinguish the difference between being popular and perceived as a mega-expert and actually being a mega-expert. She’s certainly an expert but in a room (or forum) full of experts, she doesn’t stand out, nor do any of the twitterati. She definitely does deep dives and finds nuggets like the HCU language change. I appreciate having her around.

but it feels like Google is about to drop the hammer and I'm anticipating a lot more than just low-value third-party content getting slapped.

Agreed. I always mention how Congress will pass the “Save the Children” bill that gives $1 million to the children and $500 billion to corrupt NGO’s that funnel most of it back to Congressmen. “You don’t want to save the children?!?!” This is how Google names and promotes their updates. Way more “pork” is in these updates and the true mission is never declared. It’s actively smokescreened.

Nobody should hold their breath about recovering this time around. We’ve entered a new epoch. If you recover 10% count yourself among the fortunate. And remember how exposed you are and that we all basically popped for a few years at a time before trending down. That has and will be the pattern. That’s why liquidation events are everything. Let some hopeful shmuck that can afford it hold the bag.

I keep saying how this is coming for everyone, because it’s a fundamental shift, and that cocky people with survivorship bias are peeling off by the day. One of our own said they got hit just this Saturday (not that they were cocky but they were the lucky survivor).

Any tweaks will be to double down on the current path, not to go backwards. This isn’t a mistake. This is what Google wants to achieve. They will continue to move forward with it.

I know I suddenly look like the negative nancy of BuSo lately but it’s not coming from a place like that inside me. This is simply the big picture right now and I’m always committed to ringing the alarm bell when the winds change. Hell, I started talking about this very issue in Nov 2022 when I started to smell it. That’s a full year before it officially landed. That’s all I’m trying to do now is bring awareness.

 

[chubes]

My shit has been bouncing around like crazy. It feels like it's in a good way because I'm seeing impressions on shit that hasn't seen much action for months, but its hard to tell.

 

[BakerStreet]

I have a productized service where we use content marketing to sell the service. Traffic dropped but the interesting thing is... SALES STAYED THE SAME!!!!! Sucks for you guys with Mediavine based sites but I think the HCU update will only show your webpage if it is "helpful" for the query, and "helpful" is based upon user engagement. I think it must be user engagement on the site from Chrome, Android, etc data. How else would they know what's helpful or not? LLMs can't figure out what's helpful for the user. Neither does PageRank. I think HCU is a user engagement metric.

Anyways, I'm going to re-evaluate this hunch once the update settles. Here's things I'll be investigation (food for though for you guys):

1. What is the driver of sales at my company, is it the number of blog posts I wrote?
2. What is the driver of sales at my company, is it my traffic?
3. What is the driver of sales at my company, is it the number of products I offer?
4. What is the driver of sales at my company, is it the quality and how much value my blog posts offer (ie if a few blog posts drive the sales and their quality is much better than other pages ranking for that keyword).
5. What type of keywords am I ranking for now compared to before? My hunch is that I stopped ranking for transactional and navigational keywords (or that keywords that were informational are now re-classified as navigational or transactional) and that my blog posts will only rank for keywords where the intent is absolutely informational.

Obviously, my company's not dead yet and by going through and researching 1-5, I'm going to find something. Whatever that is, that'll what I'll be scaling.

As far as my industry is concerned, the black hat guy's company is dead. 600,000 visitors to 150,000 and his team doesn't know the long tail avalanche technique. The VC backed one is fucked too. HCU turning into a webpage metric might help this one competitor but they seriously went from 3 million to 600,000 in January. IMO their drop was because info keywords were reclassifed as navigational or transactional. They have 4,000 blog posts and I doubt they'll be able to fix all those posts. Their team doesn't know how to target the long tail. They probably see 0 volume and disregard it. Sucks!

@Ryuzaki Exiting every few years out of fear of an algorithm update is only good if you run a website where the traffic comes mostly from Google. RTINGS would not exit. They built a good brand and community. Exiting is a fine solution but you lose out on future revenue if it is there. I'd just diversify so that my company's not having its eggs in one basket. Easier said than done though.

 

[Poplanu]

I've looked at her "brainstorming" sessions before to get an objective eye on my sites. @Ryuzaki do you know anyone that has used this service? If so (or not), do you have any thoughts on engaging her for this sort of brainstorming?

I know a team who contracted Haynes for an algo recovery job. They paid a lot of money (paying for the brand) and just got an over-aggressive disavow file in return that they had to manually fix up themselves. IIRC the process was not fast, either, they had to wait for quite some time. Massively overpriced for what it is was the conclusion. Grindstone has been putting out some killer vids/posts/letters recently and if I was looking for consulting I would definitely go to him first.

 

[Smith]

I know a team who contracted Haynes for an algo recovery job.

Solid feedback, thank you.

Grindtsone is definitely on my list of options.

Before engaging anyone I have some basic E-E-A-T signals to resolve (basic technically but challenging IRL) before diving into a recovery job. I'm also tracking updates following @MrMedia's suggested approach, which is showing promise at the page level.

All things considered, it would be great to see if anyone can actually pull off a recovery following an "unhelpful" classification during the September and March updates.

I personally think that any real recovery (e.g. return to pre-classification or improve on pre-classification rankings) will require a massive shock to the website that results in a complete reindexing. In short, forcing Google to reindex AND reclassify the site + individual pages AFTER all of the problem signals have been fixed.

But who fucking knows...

And just to be clear, I'm still sticking to my strategy of growing beyond Google and leveraging other channels to better serve my users/customers. But, I'm also going to leverage my site as ONE source of customer acquisition, which means trying to improve and optimize wherever possible.

 

[Encrypted]

All things considered, it would be great to see if anyone can actually pull off a recovery following an "unhelpful" classification during the September and March updates.

This!!!

I'm surprised no one has been able to pull this off. My bet is that Google made it such that no one can recover.. But assuming i'm wrong and a recovery is possible, whoever cracks this first is going to make a LOT OF MONEY!

 

[bernard]

I wouldn't count out any sites that were legitimate, someone like that fit guy website, those will bounce back if they keep working on it.

I remember Panda and Penguin and quite a few of those sites eventually recovered as links went offline and sites got upgraded and cleaned up. I had a blog that got hit and was clean some 5 years later. Suddenly it just bounced back, put content on it, ranked, sold it.

All those "me too" websites though, those started by noobs that only understood "outsource to phillippines" and nothing else, those will never recover and probably won't be online in a year.

 

[Smith]

I remember Panda and Penguin and quite a few of those sites eventually recovered as links went offline and sites got upgraded and cleaned up. I had a blog that got hit and was clean some 5 years later. Suddenly it just bounced back, put content on it, ranked, sold it.

We're you posting in the interim or did you just leave it dormant until you noticed that it was re-ranking and then started improving/publishing?

 

[Darth]

My bet is that Google made it such that no one can recover..

The problem is that Google said you can recover and as a result people have been spending MASSIVE amounts of time and money since September for absolutely no reason. It's a seriously fucxked up situation.

 

[bernard]

We're you posting in the interim or did you just leave it dormant until you noticed that it was re-ranking and then started improving/publishing?

I just let that one sit and kept adding the occassional post, mostly link swaps etc. The content was fine though, it was a Penguin slap.

I have actually de-published about 10 of my 80 posts on my Lab thread site, because I'm fairly sure I got hit with the "expired domain spam" penalty because those posts were not strictly on point. I've been redirecting every single url showing up in my analytics into relevant posts and none of those are hit.

I'd say depublishing and cleaning up links makes a lot of sense, also doing your best to make every little visual thing on your site be up to speed. Don't have empty category pages or author pages, make some custom ones. Stay clear of anything that screams low budget blog. Have a great looking footer and header. Have a nice byline and breadcrumbs. Be able to last 10 seconds without a regular person saying "this is some spam shit". And be generous with the deletion, focus on the pages that are still ranking and improve those with some unique content.

The problem being of course that if you have 1000s of posts then this will be very difficult.

 

[Zoro]

HCU Classifier is Now Page-Level Instead of Sitewide

Marie Haynes has broken down the updated language in Google's official doc's related to the Helpful Content Updates.

The key takeaway if you don't want to read it is that the HCU classifier is no longer a site-wide signal and will affect individual pages now. Which means there's a chance that people will begin to see some pages pop back up while others remain tanked.

This will allow for some actual analysis and comparison to be done finally, and possibly a lil of the ole reverse engineering.

It doesnt feel like google changed to classify on page level, atleast not only that..

See, I've got three different websites using three different writing styles – some manual, some done by AI – and they're all taking a hit. Even my mixed site, with a blend of both manual and AI content, isn't spared.

Now, if this was just about individual pages, I'd expect at least some of them to still be ranking, right? But that's not the case. It feels like there's something bigger at play here like a double or tripple classifier?

There's another thought bouncing around in my head.

Remember when you mentioned that every time Google makes a big update, they recheck everything on your site and compare it to the rest of the internet? It's like giving every post a fresh look and ranking. Maybe that's what's happening now – a sort of new evaluation process.

So, could it be that Google's doing both – looking at the whole website and looking into individual pages aswell?

 

[JOoa0ky]

My content only site got obliterated.

So I just copy/pasted word/word over to my local biz and then left it alone.

My opinion: The days of content only sites are over. It needs to provide an additional service. Adding a shopping cart, anyone? Heh.

If you're going into the energy drink niche... you should buy a bodega and then monetize the blog. Yea, the bar has been raised.

 

[Smith]

My content only site got obliterated.

I have a similar looking chart.

So I just copy/pasted word/word over to my local biz and then left it alone.

What differentiates your "content only site" from your "local biz site"?

It needs to provide an additional service. Adding a shopping cart, anyone?

I've heard a lot of people say something along these lines... But what does this actually mean in practice?

My content site is built around a service offering, with multiple services, price points, etc.

I do not have any ads on the site, I don't include any affiliate links, and I link out to other sites very sparingly.

So, what would actually signal to Google that this is site represents a legitimate business and not just a content site?

@JOoa0ky I guess I'm asking how your "local biz site" differentiates from your "content only site" besides having a shopping cart.

 

[Ryuzaki]

So, what would actually signal to Google that this is site represents a legitimate business and not just a content site?

A Google My Business profile fully fleshed out with images, hours, reviews, geo-check-ins, etc. I’ve tried literally every signal BUT that (and Apple Maps). It’s the only one you can’t fake easily. And if you pull it off you’ll get caught eventually.

The move is to approach any local business with a website and tell them you want to get them endless traffic with a new blog section and give them 5%+ revenue for their troubles. Free marketing and money.

 

[BakerStreet]

My content only site got obliterated.

So I just copy/pasted word/word over to my local biz and then left it alone.

My opinion: The days of content only sites are over. It needs to provide an additional service. Adding a shopping cart, anyone? Heh.

If you're going into the energy drink niche... you should buy a bodega and then monetize the blog. Yea, the bar has been raised.

The guys at Brighton SEO keep on talking about schemas and the important of making your brand a schema in Google's Key-value pair. The ways they know of so far are having a Wikipedia article about your brand or profile on CruchBase. I think what you show shows that having a Google My Business profile is a way to making your brand a schema too. Makes sense. Google My Business would be a data source for Google.

Everyone here knows that you need to build a brand, but how does Google know if your brand exists?...when it exists in Google's Schema key-value pair... Thanks buddy.

 

[Smith]

Google My Business profile fully fleshed out with images, hours, reviews, geo-check-ins, etc. I’ve tried literally every signal BUT that

Same. It's infuriating. I'm pissed I didn't get this set up before they instituted the "video" walk-throughs. Why can't we go back to postcards like it's 2014?

The move is to approach any local business with a website and tell them you want to get them endless traffic with a new blog section and give them 5%+ revenue for their troubles. Free marketing and money.

So you would essentially become the webmaster of their site and build out an affiliate/ad-based site on their business website? And then build niche-specific content related to that business?

It seems like there would be a lot of uncertainty and risk in building out that kind of business on the back of an existing business operation, but I guess it's one way to go about it...

if you pull it off you’ll get caught eventually

I'm not so sure about this...

If you are building a long-term business, I think the key is to actually open a short-term office... even if it's only for a week, two weeks, or a month while you're getting GMB approved.

It might sound over the top but basically, set up shop, have all your equipment and company documents on hand, and do all the video reviews etc. for the GMB approval. After you get approved, pack up and go home.

 

[Trump]

Today I've spent (wasted) a lot of time going through a site that was semi-hit. I still have some articles in top 5, but a fair bunch of them have fallen to page two, and in some cases "NOT FOUND".

The average position is 24-26, before HCU it was 3. But not all pages have been "hit", they are still bringing in traffic.

So, I made an excel sheet where I punched in user metrics for my 80 most important articles. I added stuff like average engagement time, average click per visitor, user engagement, bounce rate, time spent on page, session length, returning visitors, scrolls per visitor and so on.

Loads of different metrics from before HCU was introduced. The timespan is 8 months.

I colored each keyword based on current ranking for a better overview. For example, pos 1 to 4 is blue, 5 to 9 is orange etc.

When going through each metric, I can see 0 correlation with rankings. Really, none. Articles with very poor metrics are still in top 3, and articles with great (according to me) metrics went off the cliff.

I was hoping to see more green rows than blue or orange when filtering certain metrics, but there really is no pattern. When looking at it, it's just random. Might as well have pressed a "randomize" button in Excel.

Since this site still have keywords high up in the SERP, I wanted to know if the difference is in user behavior. From what I can see, it isn't. But I will further analyze and combined different metrics to see if I can see a correlation.

I'm also planning on adding some on-page metrics to see if I can find any clues from there.

As I wrote in a previous post, all articles follow the same structure, writing-style and it's the same niche. Imagine it's a supplement site (mine is not YMYL) where "best protein powder" and "best gainer" is still in top 3, but "best bcaa" and "best creatine" disappeared or went to page two.

Even though I am putting most of my efforts elsewhere - the investigation continues.

 

[JOoa0ky]

A Google My Business profile fully fleshed out with images, hours, reviews, geo-check-ins, etc. I’ve tried literally every signal BUT that (and Apple Maps). It’s the only one you can’t fake easily. And if you pull it off you’ll get caught eventually.

The move is to approach any local business with a website and tell them you want to get them endless traffic with a new blog section and give them 5%+ revenue for their troubles. Free marketing and money.

Precisely... hide that content behind a local biz.

That is essentially what all of the big brands are anyway, regardless of what dark past they arose from.

Kind of like how the mafia isn't completely underground. A large part of their business is but they also have an extensive front facing facade.

 

[BakerStreet]

Same. It's infuriating. I'm pissed I didn't get this set up before they instituted the "video" walk-throughs. Why can't we go back to postcards like it's 2014?

I think you can just show them documentation and your WP backend. From https://support.google.com/business...f-my-business-doesnt-have-a-physical-location

What if my business doesn’t have a physical location?
You can still record the location where you offer services, the equipment you use, and their documentation associated with your business.

• Examples of location:
  • A street name and number on the building.
  • Multiple store signs around the location.
• Examples of equipment:
  • A branded vehicle or marketing materials.
  • Your workspace, tools and equipment you use to serve your customers.
• Examples of documentation:
  • Business registration, invoices, utility bills for the business.
  • Other documents with your business name.

 

[Politico]

When looking at it, it's just random. Might as well have pressed a "randomize" button in Excel.

I keep saying it but no one wants to hear it - GOOGLE IS FUCKIN WITH EVERYONE and they always have been.

The nature of every single update since the beginning of my time in SEO is RANDOMNESS.

Random shit sticks, random shit gets hit.

I'm more and more starting to think Google doesn't know what it's doing beyond it's original PageRank algorithm. Links are the BEST way to determine relevance and authority.

Edit 3: And then they blow up the serps at regular intervals to keep us guessing and chasing our tails.

Document "quality" is not something they can programatically define. They try to fake it with user metrics, but obviously the data from OP shows they're not that great at that either.

You Churn. It Burns. So churn some more. Or build a business that's not solely reliant on search traffic.

If you want to play in SEO, get comfortable with losing all your work and having to start again.

Edit: There's a reason SEO has always been seen as the super spammy BHW style of marketing. Because it is, at it's most-effective, spamming the internet.

That's how it's been since the beginning, and especially so since Panda-geddon.

NO ONE has authentically reverse-engineered an update.

NO ONE has any fuckin clue why Google does what it does when it hits sites.

NO ONE has any idea of why some recovery efforts work and others dont.

You know what we do know, and have known for 20 years?

It all seems.... RANDOM.

Edit 2: This is a FEATURE, NOT A BUG

/rant

 

[bernard]

That local biz method will probably work for a while, but I feel as if Google will simply classify your biz as a content site then.

I don't think the issue is content as such, but the type of content, the hyperoptimized content.

If you just write a bunch of content and never think about SEO, I strongly doubt you'll ever get punished.

There are also other methods you can do with content sites, but it's something imo. that should be considered long before getting hit by the HCU.

You want to be a niche hub, not a content site. That means stuff like having:

• Job board
• Forum
• List of service providers
• List of geographical areas of interest
• Bespoke niche services (consulting, teaching etc)
• Activism (political, local, niche)
• Sponsorships
• Surveys
• Activities and contests
• etc

I mean, these are the things that a legit content hub will have. It's also fairly easy to get involved with.