Google Updates and SERP Changes - July 2017

[Versado]

---

System: The following 4 messages were cut out of thread at: https://www.webmasterworld.com/google/4851947.htm [webmasterworld.com] by robert_charlton - 6:51 pm on Jul 1, 2017 (PDT -8)

---

Seeing a lot of changes even on a daily basis specially with new URLs and pages that have got new backlinks. Some pages appear on top 5 and after a few hours they return back to page 2.

On the other hand, I haven't seen any movement on pages that have been already ranking on top 3 for months.

It seems to me like an algo refresh not related with Panda or Penguin which has not finished yet.

[MayankParmar]

My WordPress site was hacked, unwanted articles were published. Will this affect the site's SEO? Although, articles have been removed and it had no links to any site.

[SEchecker]

Our website traffic drops at night times (11pm - 4am) on under 5% in last 14 days. No changes where done on website from our end.

Former situation was that most of conversations and sales where made in nigth times. Now, lower conversations and less sales over day times.

Website language "english" target location is "not listed" server location DE but most traffic from US.

[reseller]

AccuRanker, RankRanger and SEMrush Sensor are showing high level of SERPs fluctuations on July 9, 2017 . Maybe the start of a new algorithm update, who knows :)

[Martin Ice Web]

Next goonopdate. google forcing big brands to the top. Content doesn´t rule anymore. Its links and money.
From users view organics are full with very expensive shops, but on the right hand you have google shopping results. Very clever.

[MayankParmar]

Was there another update?

[MayankParmar]

I agree with you, Martin. Content is no longer the king. Bing is best <3

[reseller]

A relevant post from Search Engine Roundtable of Jul 10, 2017 :

Signs Of A Google Search Algorithm Update Over Weekend

There are some early signs of a Google search algorithm update over the weekend, around Saturday and Sunday. The chatter is still very quiet but it was the weekend and that might kick up over the next day or so. Plus, pretty much all the tools are on fire, showing huge volatility around June 9th.

[seroundtable.com...]

[martinibuster]

Barry "failed to find any patterns" after analyzing 70 sites...

The reason Barry couldn't find anything is because the traditional way of diagnosing changes to Google's algorithm is no longer appropriate. The traditional approach is to identify what SEO Mistakes Google is Targeting. But Google's algorithm is an Information Retrieval algorithm, not a 200 Rank Factor Targeting Algorithm. Thus it's not surprising that Barry didn't find patterns and Barry should be commended for not reading a false positive into the results (something that others have done in the past).

The algorithm is designed to find the appropriate answer. Rank Ranger can't show you how Google is understanding search queries better, that's why Barry didn't see anything.

This longstanding approach by the SEO Community that is focused on how Google is targeting bad sites, bad quality, excessive keywords is a habit that needs to stop because it's keeping the SEO industry from seeing what is really going on.

That "targeting" mindset approach is what led to all the mistaken theories surrounding the so-called Phantom Updates that were 100% wrong.

Google's algorithm is not a Spam Targeting Algorithm, it is not primarily hunting down bad quality sites to demote. The algorithm is not ticking boxes on the 200+ Ranking Factors list. Seriously, some of you still believe in 200 Ranking Factors? I wrote about this on my blog and on WebmasterWorld a few months ago.

SEOs need to stop thinking in terms of 200+ ranking factors. I wrote about that on my blog and here months ago. Once you do that THEN you can begin to understand what Google is doing. Here's a hint: It's called Information Retrieval.

[edited by: martinibuster at 3:40 pm (utc) on Jul 10, 2017]

[londrum]

how do you know? no one knows. I'm not even sure that google employees know anymore -- their algorithm is probably so complicated and multi-layered that it's impossible to predict.

[martinibuster]

how do you know?

I know because I research the algorithms and see how they line up with the official information out of Google.

Think I care whether you or anyone else believes me? Feel free to ignore what I write. I encourage you to ignore me. ;)

Good luck,

Roger Montti

[reseller]

@martinibuster

Thanks for your time and content-rich contributions. Much appreciated :)

[Shaddows]

Sure, but the psychology is important.

Every webmaster believes their site is the best (otherwise they would improve it, presumably). So, a demotion must be a penalty. Loss of traffic is a penalty. Everything is a penalty.

And that penalty is a false positive.

So, Step 1: Find what Google was penalising
Step 2: Complain that isn't what your site...
...I mean...
Step 2: Fix the trigger for the false positive.

Not saying @RustyBrick is part of this, but there is a whole industry catering to this world view. Sales 101- sell what the customer is buying.

[Totalx]

@martinibuster

I read your article, and it was interesting. Can you recommend who or what else to read on this subject?

[seoskunk]

OK Shaddows

How would you suggest a webmaster deals with this situation

Step 1: Find what Google was penalising

Answer: Offsite factors beyond the webmasters control

@martinbuster Interesting, you have triggered some thoughts for sure :)

[martinibuster]

Can you recommend who or what else to read on this subject?

Off the top of my head, I can only think of Bill Slawski (Google the name). Bill does an excellent job of documenting patents and leaves it to you to connect the dots, he doesn't do that for you (which is reasonable).

I don't recommend reading articles by those who see nearly every change at Google in terms of "quality updates" because they are biased to seeing everything as the same thing- which is obviously wrong.

I do not recommend anyone who relies on correlation studies, either. What they do is massive studies of sites that rank and note various on-page and off-page factors then come to conclusions about the correlations of qualities of sites that tend to rank: Their conclusions are always wrong, 100%.

For example, some SEOs look at anchor text percentages of sites that tend to rank then use that data to create a formula for the ideal anchor text ratio. Their idea is that sites that a certain anchor text ratio correlates to high ranking sites. But it's a 100% wrong conclusion because you can do the same thing with the same sites for CMS and find that sites that use WordPress correlate to high rankings- which would also be a 100% wrong conclusion.

Run from any so-called SEO who publishes a correlation study. If they are unaware that correlation studies are 100% fake (see this site [tylervigen.com]) then they do not know SEO. Period.

[kirtisharma]

Is there any new update on 9th july? I feel some fluctuation in my Ranking. I am surprised if there any google update then why not google confirm it.

[Shaddows]

correlation studies are 100% fake

Erm, possibly for SEO. Possibly. But I would go for a smaller figure than 100%

Yes, you can misuse or misconstrue correlation. You can get small datasets to correlate, or just pick two datasets because the already correlate.

That doesn't mean all correlation is spurious. Machine learning would not work without some veracity in statistical analysis.

why not google confirm it.

Google became unreliable at confirming updates when Matt Cutts left. Well, they started us unreliable and now they are somewhere between inscrutable and deceptive.

Anyway, I don't think Google actually does many overt changes these days, just rafts and rafts of minor tweaks, including led by RankBrain (which uses statistical analysis)

[MayankParmar]

I can confirm there was an update. I'm hit by it again, lol. Now my views from Google is 1500-2000 after the latest update, 60% dropped. A month ago, it was over 10K :(. Before Fred it was 20K.. now nothing is left :(.

[Choiceed]

thanks for answers. I discovered that in different regions my position is very variable. google search console tells me, that my position hasn't changed at all.
it is no something new, but it is so difficult to get some gain

[martinibuster]

correlation studies are 100% fake

Erm, possibly for SEO.

Not possibly. It's a 100% certainty that all correlation studies of the SERPs result in wrong assumptions. There are literally thousands of factors that go beyond the 200+ ranking factors in order to produce the SERPs. Nobody will obtain meaningful or useful information from conducting correlation studies of a single factor in isolation or even 200 factors in isolation. There are many factors outside of the 200 Ranking Factors that affect why pages are ranking.

An additional problem is that some data is being twisted to fit the Traditional Rote SEO Worldview. For example, an often cited study about heading tags in the SERPs is actually a misrepresentation of the facts. Yes, heading tags appear on many sites that rank in the top 100 but it's also a fact that Heading Tags are commonly used for styling purposes. Duh.

Duane Forrester, formerly of Bing, posted a fantastic article today about how modern search rankings have gone beyond 200+ ranking factors. [searchenginejournal.com] It's great to see someone else talking about this to get SEOs out of the 200 Ranking Factors mentality.

Here's what I posted about his article on Duane's Facebook Feed:

That article is spot on. I've been trying to communicate this for the past few years but some SEOs are too hung up on their 200 Ranking Factors to fully wrap their minds around it. When some of them try, about two paragraphs into it they revert back to their traditional 200 Ranking Factors SEO.

So I see this as a contingent of SEOs who cling to old traditional methods rooted in the past versus how search engines have evolved way past anchor text and keyword pattern matching, which is the present. 200 Ranking Factors SEO versus Modern SEO that has more dimensions to it.

The sad thing about this is when SEOs use traditional SEO ranking metrics to understand pages that are ranking because of non-traditional ranking factors. <-- that leads to false conclusions such as "they're ranking with footer links" or that the ranks are powered because of EMD effect. The SEO industry is rushing to red herrings to understand why some pages are ranking, to their own detriment.

[MayankParmar]

@Choiceed, check the impressions. It would be down. My position improved but impressions are down. Maybe because of those big boxes article or whatever Google calls it.

[samwest]

I can confirm there was an update. I'm hit by it again, lol. Now my views from Google is 1500-2000 after the latest update, 60% dropped. A month ago, it was over 10K :(. Before Fred it was 20K.. now nothing is left :(.

Yep, I would concur...something big hit again and traffic is all but gone. Amazingly, Google keeps sending insight reports saying traffic is up hundreds of not thousands of percent on many pages. Regardless, sessions keep dropping. Must be the weather....right. Site used to get 5000 sessions per day, but now at about 100 and still on page one for many terms. What does that say about Google SERP's?

Just glad I found a job when I did. The web is dead to me.

[Shaddows]

Duane Forrester, formerly of Bing, posted a fantastic article today...

That article is interesting. @iamlost had a thread on here recently talking about the use/misuse of the word "brand" [webmasterworld.com] - it mostly covers similar ground.

I see what you mean about correlation studies to extract information about a specific factor, but I'm not sure that is what Barry intended. In the past, Barry (not always successfully) tried to find common elements, not just perceived ranking factors. Finding an explanation-in-common for multiple outcomes seems to be a valid method, as I found this quote in an excellent article from April...

A reasonable person can assume that it could very well have been an improvement to Google’s ability to understand user queries and/or web page documents (probably both!). A change on this order would affect sites with poor links as well as ad heavy sites and pages with thin content. Mystery solved? Probably!
[Emphasis theirs]
Assume attribution allowed! [martinibuster.com]

In a somewhat related vein, Has anyone noticed that for Brand/Nav searches, it is now extremely common, at least here in the UK, to find below the subject site:
- At least 1 review site
- At least 1 SM site (FB or Twitter, with increasing frequency LI)
- Frequently a "Company Valuation" site sourced from public records (companies house in UK- frequently companies house is the actual link)

Almost as if Google thinks those are important things a searcher might rely on.

[reseller]

@Shaddows

why not google confirm it.

Google became unreliable at confirming updates when Matt Cutts left. Well, they started us unreliable and now they are somewhere between inscrutable and deceptive.

Well said :)

[martinibuster]

I'm not sure that is what Barry intended. In the past, Barry (not always successfully) tried to find common elements, not just perceived ranking factors.

Barry's approach is perfectly valid and recommended.

However a pitfall in that approach is bias. This happened when some SEOs in the industry (not Barry) ignored contrary evidence because it didn't fit the working theory of what the update was "targeting," as happened with the so-called Phantom/Fred updates.

Additionally, the approach of finding common elements of sites that lost positions will not catch situations where Google changed how it interprets relevance. In which case the sites losing positions were not penalized, not targeted, and not demoted because of a lack of quality.

Thus, the reasons why those sites lost positions will not be discovered by the techniques and tools SEOs commonly use in their site reviews. No tool will be able to diagnose that kind of algorithm change. I have diagnosed ranking problems by identifying the issue outside of the site itself, then applied the solution to help the site regain rankings.

Good luck,
;)

Roger Montti

[MayankParmar]

There are some website which has the articles with 50-70 words. 90% of the words are actually the keywords written in all possible ways. It ranks 2nd or 3rd, while I'm on 2nd page with high quality (500+) words article. Lol.

[Lara Alexandor]

Dear Experts,

I need your expert advice on why Google is deindexing pages of my website whereas all pages are unique in meta's, url, heading and also there is a variable change in content as well. All these pages are being created with a php script and generate run time. I have submitted all pages in Google through sitemap files and the indexing of these pages was increasing day by day. But now suddenly indexing is getting worse day by day and website traffic is also effected.

The indexed pages in Google were almost 1 million and now these are only 500.

I need expert advice about why Google is doing this and how i can increase indexing of pages again?

[edited by: engine at 12:24 pm (utc) on Jul 13, 2017]
[edit reason] No specific sites, thanks [/edit]

[rustybrick]

Truth is, when there is no clear and obvious relationship between the sites that were hit (at least the sample given to me), then we call it a "quality update." :)

But I think that is your point.

Thanks everyone.

[Martin Ice Web]

@Lara, my guess is that g is testing the mobile index live. I think the mobile index is not far as big as the desktop index.

@rustybrick,

there is no clear and obvious relationship between the sites that were hit

maybe you should not look at the sites that been hit but on the sites that now moved up. In my vertical it is big brands. So i wouldn´t call it quality update but likely a brands update.

Nevertheless u do a good work. Go ahead. In the end only google knows about their updates.

[samwest]

Pinterest is taking up WAY too much space in the serps, almost as bad as "e-zine articles" used to be. BTW - if you remember that site, you are a veteran...they are located just down the road and their big fancy building is up for sale. No surprise.

While Pinterest may be the flavor of the month, they are decaying in quality fast. I found a ton of my stuff indexed by Google on Pinterest that was scraped and pinned. What a mess. Every time I find one, I send feedback to Google telling them they can send $500k to my PayPal account for damages. lol