Google Core Update May 23 - June 3, 2021

Google *finally* confirms 11 days later.

[TalkativeEditorial]

Yower, SEMRush is freaking out. A few of the other sensors seem to be spiking too.
This is hell. Never know what to do anymore, where to look...or what is actually evening happening.

[TalkativeEditorial]

Well part two of the June core update should start rolling out in the coming days. So this might be the first signs of that.

[yollo03]

No it's not. They said they will announce all core updates when they roll them out. This is part II of the spam update, the same thing happened last time. It took a few days for the sensors to pick it up.

With core updates semrush is well above 8 and this is for all topics.

Edit: When the core update will be out, don't count on it to do anything in your favor. There is a greater chance it will do further damage. So those that were hit (like me) by December core update, be mentally prepared that it may make things worse.

[mzb44]

My site got a super massive boost the last 2 days that almost completely reversed the actions of the December core update.

Now, today, this has been completely reversed and it's back exactly to where it was before around two and a half days ago.

You can see on the analytics charts precisely to the hour when it started and when it stoped. It's like a switch was flipped on and then off again.

I think they may be testing the next core update.

Or it was the spam update which was then reversed.

[yollo03]

The spam update will hit my website tomorrow (most likely). It usually happens with core updates, some sites get a temporary boost or decline. I experienced it once, it lasted for less than 24 hours. The sales I received on that day were equal to almost a full month at the time.

[ichthyous]

@mzb44 I saw this on Monday...almost normal traffic and my first conversions in a week. Yesterday traffic stopped at 9am sharp and remained very low all day. I hope that isn't a portent of what is to come. Today I wake up to a big overnight surge which dropped off every hour to zero at 9am. This pattern is clear for many of us here...the traffic is simply vanishing for the prime hours for business.

There has been a very big drop in traffic classified as "direct" to my site since the weekend. We're talking down by 40-60% daily, with a reversion back to normal on Monday, then back down on Tuesday. This is mostly traffic from Google Images...has anyone else seen this drop?

[TalkativeEditorial]

That "something went wrong" error seems to have returned when requesting indexing in GSC.
Tried across a few different sites - anyone else stuck also?

[yollo03]

It happened earlier but it seems to be ok now. Maybe they are working on a bug they discovered.

[mzb44]

Edit: When the core update will be out, don't count on it to do anything in your favor. There is a greater chance it will do further damage. So those that were hit (like me) by December core update, be mentally prepared that it may make things worse.

I completely agree here.

Historically, if you were hit before the chances of recovery were against you, even if you did everything in your power to fix or improve things.

This reminds me of something interesting I saw this week. I'm seeing several sites that were launched after the May 2020 core update - the one that hit me massively - now all outranking me by A LOT.

The interesting thing about these sites is that they aren't that much different from my site or other sites existing from before May 2020. In a lot of cases those new sites are almost copies of my or other older sites. Not scraped and I wouldn't call them spam necessarily but they clearly got "inspiration" from all the older sites and essentially rewrote whatever those sites did and that's that (I recognise their writers from Upwork - all are from "cheap" countries).

One site even copied my design and structure. Different branding, colours etc. but the site structure is very obviously inspired from my site. This site now gets approximately 100x more traffic than what I get, even though I'm better in every way (expert content creators, perfect UX, ultra fast speed, super modern design, etc etc and all that...)

So, this made me think: What if the May 2020 core update had a component that was ran only that once and whatever it did it still affects all the hit sites existing from before May 2020 but any new site launched after that date remains unaffected. This would explain all the reports of newer sites outranking older sites mentioned here, even with inferior content and lower stats.

I'm particularly noticing that it's sites launched post May 2020 that rank very high and are usually outranking older sites.

I don't particularly have anything against all these new sites that now outrank and copy me. But it's incredible that in 2021 we still have Google updates (or components of larger core updates) that apparently are only run once a year or even less frequently (or maybe just once and never again?) and if you get affected by that you get stuck, while new sites that essentially copy everything you did (and do it less competently) meanwhile rank extremely high because they did not exist yet when that update was run.

What if whatever they did in May 2020 / December 2020 won't be done again? We all just remain stuck forever? Should we just cut our losses and create new sites instead?

At least in my case, looking at the copycat sites outranking me, I'm 100% convinced that if after May 2020 I instead would have created a completely new version of my hit site - similar content, similar everything - it would now rank MUCH MUCH higher that my older site, even though I would be almost exactly the same site.

Is Google incentivising churn and burn instead of incentivising the improving of existing sites? - In this particular scenario this is my takeaway.

Edit:

So, the point I'm trying to make is that if your site gets hit by a core update, you can essentially re-create your old site almost identically as a new site and this new site will now rank just fine for years, while the old site can remain stuck for years.

So it makes more sense to keep churning out new sites.

These are perverse incentives.

2020 and 2021 are now the years with perhaps the most spam in Google ever and largest number of spam sites and pages created.

I'm sure it's completely unrelated. :-)

[edited by: mzb44 at 2:46 pm (utc) on Jun 30, 2021]

[yollo03]

Some high authority websites are abusing their ranking and publish new posts with less than 250 words, knowing very well they are likely to enjoy very high ranking regardless of the content. It's happening in my niche. They publish dozens of posts per day, all less than 250 words each.

About May 2020, it makes sense in a way. If you get penalized for something they released in May new websites can't be penalized for it because they never existed. The interesting question is whether they will do it again, like once a year or once every 5 years?

At some stage people may come to the realization that there is no point in watering dead plants. I will probably launch a new site after the next core update and pay 80% of my attention on that and 20% on the existing website.

[mzb44]

Some high authority websites are abusing their ranking and publish new posts with less than 250 words, knowing very well they are likely to enjoy very high ranking regardless of the content. It's happening in my niche. They publish dozens of posts per day, all less than 250 words each.

Oh yeah for sure. I've seen this many times. This is because of the authority vs. relevancy adjustments.

This is a transcript from the Google video "Trillions of Questions, No Easy Answers" (recommended watching!): "Every query has some notion of relevance and some notion of quality. We’re constantly trading off which set of results balances these two the best. [...] We have long recognized that there is a certain class of queries like medical queries or finance queries. In all of these cases, authoritative resources are incredibly important. And so, we emphasize expertise over relevance in those cases. We try to get you results from authoritative sources in a more significant way."

This is now clearly expanded broadly over the entirety of search and not just medical/financial. Expertise/authority > relevancy.

About May 2020, it makes sense in a way. If you get penalized for something they released in May new websites can't be penalized for it because they never existed. The interesting question is whether they will do it again, like once a year or once every 5 years?

Yeah exactly. To me it clearly seems that at least one or more components that were launched by the May 2020 update have not been done afterwards anymore in subsequent core updates. The fact that new sites launched after May 2020 that are lower quality copycats of the older sites now outrank the older hit sites and never gotten themselves hit by new core updates shows that some things done in May were not (so far) done anymore in the subsequent core updates.

Agreed that if at one point you see you aren't recovering but new - and similar - sites launched after the update that hit you rank just fine, then it's perhaps a better idea to cut your losses and just create a new site. For all you know, whatever hit you that time will never be run again and you will never recover no matter what you do.

[Samsam1978]

I cannot even face the thought of having to start again. It took me over a decade of work which is going down the pan since the May update. Copy cat (poorly written) websites and news sites are outranking me with little content, they even use my youtube videos!
Yesterday I got 40 people email saying how they love what I do. Why has this update done this, it is not following what the users want, using some ranking AI that is allowing all this spam to get in and scrape our old sites, messed up poorly written content outranking old websites with authority.

My traffic is 40% down from January 2020

[edited by: Samsam1978 at 3:21 pm (utc) on Jun 30, 2021]

[mzb44]

I cannot even face the thought of having to start again. It took me a decade of work which is going down the pan since May update. Copy cat (poorly written) websites and news sites are outranking me with little content, they even use my youtube videos!

Eventually we need to face it.

If we won't recover despite improvements and if copycat inferior sites keep continuing to rank/outrank, then we must entertain the possibility that whatever hit us in May 2020 was not/won't be refreshed or run again.

At one point in time it will not economically make sense to keep working on these sites, even if sometime in the future they might still recover.

At one point in time it will financialy make more sense to start a new site.

[ichthyous]

I cannot even face the thought of having to start again. It took me a decade of work which is going down the pan since May update.

Try 18 years. Nothing lasts forever, and I am having to face the post-Google era as well. It probably won't be as profitable that's for sure...will have to spend a lot more for eyeballs. I am rethinking every possible way to break my dependence on organic search, but it will all take a lot of effort and years to recover. From January - March 2021 I was bringing in money hand over fist from search alone, so I resent this train wreck since mid-March.

@Samsam Have you tried spending time on new links? As many as you can possibly get. I have made a lot of on page changes and added content, but it hasn't helped much.

[westcoast]

This reminds me of something interesting I saw this week. I'm seeing several sites that were launched after the May 2020 core update - the one that hit me massively - now all outranking me by A LOT.

The interesting thing about these sites is that they aren't that much different from my site or other sites existing from before May 2020. In a lot of cases those new sites are almost copies of my or other older sites. Not scraped and I wouldn't call them spam necessarily but they clearly got "inspiration" from all the older sites and essentially rewrote whatever those sites did and that's that (I recognise their writers from Upwork - all are from "cheap" countries).

I'm seeing that in our vertical too. New sites just dominating anything old that is not one of the top-50 large web properties. Our site was one of the earliest on the web (established 1994!), and has seen a slow downward drop year after year for over 10 years. Compared to our competitors we still offer far more and better content, and have stayed current. It feels like no matter what we do we just can't escape the gravity pull of time.

I don't buy that Google is intentionally trashing old websites in favor of new ones. I think the more likely explanation is that there is a major bug in the way that google processes / handles old websites (outside of those few truly massive old players).

Think about it... the size and complexity of a data set can result in HUGE differences in algorithmic conclusions. I suspect what we're seeing is the action of algorithms coded to work best on perhaps 0 to 10 years of data now choking on sites with 25 years of legacy data. Algorithms tend to have very different output when you get massive amounts of legacy/historical data thrown at them, vs small data sets. I suspect the algorithms are looking at the smaller data sets of new websites and going "ok, we know how to deal with this. we always have".

However, when the algorithms operate on 25 years of data, backlink profiles polluted by 25 years of spam and content scraping and noise, site structures polluted by 25 years of redesigns and historical 301s that Google *still remembers even though they have not been there or used in over a decade*, it totally makes sense that they will produce erroneous rankings for these older sites. Data from new sites is just so much cleaner and less noisy.

The reason I bring up this idea of "unintended old site penalty" is that we had a similar issue on our site when producing rankings for data sets ourselves! When our algorithms operate on newer data sets, we get far more variance and thus a greater chance of outliers to both the upside and downside. However, with the entities with long histories and massive amounts of data came a reversion to the mean, and so a lot of older data sets which should rank higher were dragged down by the statistical tendency of reversion. I'm fairly sure this idea of "large, old data sets" creating ranking anomalies must be a pretty standard outcome in statistical studies.

Anyway, outside of the other very real issues of Google pushing their own stuff, my contention is that Google just isn't treating old websites with a lot of historical crawl & backlink data fairly, completely unintentionally. The few large sites like Amazon and Pinterest escape that because of their overwhelming authority, but large, old, mid-sized sites simply get sucked into a statistical vortex.

My gut is that backlinks are the root problem. So many old websites have a large % of their total backlinks as spam or low quality. From an algorithm's perspective an old site's link profile often looks like a tire fire. How does an algorithm rank that fairly vs a new site with a tiny number of good quality links and nothing else? My guess is it doesn't.

[ichthyous]

I don't buy that Google is intentionally trashing old websites in favor of new ones. I think the more likely explanation is that there is a major bug in the way that google processes / handles old websites (outside of those few truly massive old players).

I'm not sure if Google is favoring new sites...perhaps fresh content, perhaps discounting links, and I have really been wondering if sites with too many 301 redirects are being punished. I deleted a huge number of old redirects for pages that never got much traffic back in 2016 when I redesigned my site, but Google still comes looking for them so I had to replace them. I saw no boost in placement for having cut my redirects almost in half, but it did improve my server response time.

My gut is that backlinks are the root problem. So many old websites have a large % of their total backlinks as spam or low quality.

Something major happened regarding links, but a lot of newly ranking competitors have also accumulated junk links. I have seen one and only one of my small independent competitor sites skyrocket in the last few months and beat out large corp sites. He's been around for 15 years.

[christianz]

Are they intentionally pushing down salt-of-the-Internet niche champions to dumb down SERPs and encourage interaction with googlespam and boost priority publishers or have they failed spectacularly with a brand new AI-based ranking model, supported by AI kool-aid drinking top managers at the search team?

A - malicious intent or B - fundamental mistake? I still can't decide.

[TalkativeEditorial]

Is Analytics broken now too?

[universenet]

Is Analytics broken now too?

seems yes

[Markedd]

Almost had a heart attack when I saw zero in Analytics, but could still access my website.

[joshd2]

Think about it... the size and complexity of a data set can result in HUGE differences in algorithmic conclusions. I suspect what we're seeing is the action of algorithms coded to work best on perhaps 0 to 10 years of data now choking on sites with 25 years of legacy data. Algorithms tend to have very different output when you get massive amounts of legacy/historical data thrown at them, vs small data sets. I suspect the algorithms are looking at the smaller data sets of new websites and going "ok, we know how to deal with this. we always have".

However, when the algorithms operate on 25 years of data, backlink profiles polluted by 25 years of spam and content scraping and noise, site structures polluted by 25 years of redesigns and historical 301s that Google *still remembers even though they have not been there or used in over a decade*, it totally makes sense that they will produce erroneous rankings for these older sites. Data from new sites is just so much cleaner and less noisy.

This sums it up perfectly. This a "bug" in the algorithm that has been occurring on and off for the last year.

I believe it also correlates to Webmaster Tools showing "Indexed, not submitted in sitemap" instead of "Submitted and indexed". The algorithm stops understanding how to index some large older sites when there is an update

[TalkativeEditorial]

Almost had a heart attack when I saw zero in Analytics, but could still access my website.

Same. I mean it's been pretty bad over the last few months but saw that and thought the absolute worst. I hate this.

[universenet]

I believe it also correlates to Webmaster Tools showing "Indexed, not submitted in sitemap" instead of "Submitted and indexed". The algorithm stops understanding how to index some large older sites when there is an update

if you see some disorder so this is normal, disorder will be minimum to september
In september google will get comeptitions search engines in EU zone and after in other part in world

[yollo03]

My analytics dont work either BUT, if you inspect a URL in the console, the analytics will show a visit to that url you inspected. So it is not completely broken. Quite odd...

Edit: I am assuming it was shut off intentionally rather than a bug. Who knows what they are planning?

[MayankParmar]

Analytics is having issues.

[yollo03]

Analytics is working again.

Edit: I might be imagining but I think analytics is a bit faster than it used to be.

[mzb44]

Are they intentionally pushing down salt-of-the-Internet niche champions to dumb down SERPs and encourage interaction with googlespam and boost priority publishers or have they failed spectacularly with a brand new AI-based ranking model, supported by AI kool-aid drinking top managers at the search team?

A - malicious intent or B - fundamental mistake? I still can't decide.

I honestly think it's probably both.

[RedBar]

Has someone switched off T'Internet today, 1st July?

Incredibly dead across all my sites, I actually had to check my server was not down, very little spam and few emails too!

[RedBar]

Just going through my logs and the switch for me was flipped about 23.00 UK time Weds night.

Anyone else?

[RedBar]

More info, 90+% of visitors since 23.00 have been single page views, very unusual and since 06.00 UK time this morning zero USA visitors whatsoever, that is most unlikely at 22.00 West Coast time, nothing for 5 hours?!?!

[hemanath]

Thanks for the useful information. I hope you will share some more content. Please keep sharing!