Google Updates and SERP Changes - April 2019

[dollarsound]

The following 2 messages were cut out of thread at: https://www.webmasterworld.com/google/4937425.htm [webmasterworld.com] by robert_charlton - 11:52 pm on Apr 1, 2019 (PDT -8)



Today traffic is 35% up. This site wasn't affected by these previous updates in march, in august had an increase in organic traffic also, and since then has been fluctuating

[edited by: Robert_Charlton at 7:58 am (utc) on Apr 2, 2019]
[edit reason] Cleanup after thread split to new month [/edit]

[swright]

Anyone notice from time to time direct traffic shooting up while Google traffic drops for a while? I think it's some kind of mislabeling that happens for some reason when the people at G are doing tweaks.

[Robino]

Yes. I came to this thread to see if people were seeing the direct traffic fluctuations.

For my ecomm. sites it's not so much a direct traffic increase this time. But transactions attributed to direct traffic are up 70% while organic is down 30%.

[Appstar Solution]

looks like there was another update or tweak again today.
many page returned in index but not ranking in top 100.

[glitterball]

One change I've seen over the last few weeks is the return of a competitor's website that hasn't been updated in 10 years, is not mobile friendly (couldn't be more mobile-unfriendly) and is not https. Has the mobile-first policy been relaxed?

[Erwitt]

My site is indexed again in GCS, but traffic is still 30% lower. I'm really worry about that...

[Milchan]

Anyone notice from time to time direct traffic shooting up while Google traffic drops for a while? I think it's some kind of mislabeling that happens for some reason when the people at G are doing tweaks

In general what seems to happen after an update is that things are unstable and most/many experience these ups and downs, traffic spikes etc. I think it is basically down to the fact that it takes a while to rollout out and update across the millions of nodes google has plus the update algo itself has to rerun across the index , perform calculations, adjustments and apply the AI again which all takes time. I.e its not just a case of apply an update and there you go, the google search is complex real time algo that Id guess goes through a number of passes to reach its target state - that would of course result in shuffling and spikes.

Its all a guess of course but its based on what I have observed this last year or two when ever there has been a change. Usually there is an update, lots of movement,ups and downs, then just as it seems like it might settled they apply an adjustment or tweak and we get movement for more days, settle again then a week or two later another adjustment and we get thrown around again before we finally get a month or so of relative calm before it happens all again.

Of course different sites / niches are effected to varying degrees each time as google ensures that as many of us are nervous wrecks as possible.What can we do? You can try to make as good a site as possible and follow google guidelines but Im really not sure it makes a difference anymore to be honest. Maybe if you have any money left invest it in either google shares or a pharmaceutical company that makes anxiety meds as these are the best bets as far as I can see.

[ichthyous]

@swright Yes my direct traffic has also skyrocketed and it has a very high bounce rate too

[broccoli]

I’ve just discovered (not for the first time) that Google is using some copyrighted custom made images from my site to advertise someone else’s site in featured snippets. And it’s a god-awful site. This is infuriating. Does anyone know if there’s a way to prevent this from happening?

[azlinda]

Why can't they just be a search engine and forget all this constant updating and tweaking!

[seomotionz]

@ichthyous We also got some direct traffic yesterday. But after doing some digging we found that it was bot traffic from a unheard town of Brazil.

[StoneSolid]

Because they don't know what they are doing, and I mean that in a literal way.

Genius engineers that made the initial set of ranking algorithms are long gone, and they left millions of lines of ranking code that was patched by thousands of other engineers through last 20 years.

New engineers can't fix old code so they just add patches on top of existing and that is why there are random bugs, when rules collide against one another.
That is why you get serious spam sites popping back up on top, de-indexing pages bug, and of course, most importantly - that is why sites get insane boosts and drops after each update.

It is so obvious really.

Why can't they just be a search engine and forget all this constant updating and tweaking!

[Robert Charlton]

It is so obvious really.

Stonesolid.... I don't think it's so obvious at all.

To me, and I'm conjecturing a lot... Google has prior states of its index pretty carefully archived, with "restore points", very carefully controlled by routines and databases that do nothing but version control.

The "bug" might have been in some sort of recursive process, perhaps a filter that Google instituted to prevent spammers from, say, using the add url function excessively. Perhaps the bug was somewhere in a stage that was further back than Google expected... and Google has had to unwind more layers than expected. These engineers are as bright as are available... they go through rigorous internal training and establish their own levels of qualification, which go a lot higher than you and I do, but maybe I should speak just for myself... as you sound like you believe you've got it figured it.

No doubt they've overreached here and missed something, but I'm thinking that their accomplishments are impressive enough that I'm less quick to be dismissive. YMMV.

[Milchan]

Why can't they just be a search engine and forget all this constant updating and tweaking!

I get the desire for this but of course they are going to try to "improve" which requires updating and tweaking - of course many people will disagree as to if some things are improvements or not (myself and many on here included of course).
Ultimately though, Google is out to make money for Google, to consolidate and further its power and control. They do not care about us nor do they care about the users feelings. They would only react to user dissatisfaction if that resulted in an effect on its bottom line or effected its growth negatively.

Re the engineers, I don;t agree that it is a case of the best engineers that created algos years ago having left. Google will have more and even better engineers now but the simple fact is that now google is a monolitic beast with huge teams of developers commiting over 16000 changes per day to their products. They have sophisticated tests within their CI/CD pipelines to try to catch any errors, use a deployment / container management system they developed called Borg (which is the predecessor to Kubenetes - one of the fastest growing technoglies in the SysOps / DevOps world) that enables them to deploy, an rollback changes on the massive scale required.
My point is , with all this going on (do a search and read up on Borg and Site Reliability Engineering if you want to learn about how google manages to administrator is infrastructure - it is on a scale that is hard to fathom) there are going to be the odd glitch that gets through every now and again and that certainly doesnt mean the engineers dont know what they are doing.
I would concur though. along the "dont know what they are doing" line in relation to their use of AI and by that I mean , they dont know what the AI is going to do exactly. By the very fact that AI must have a degree of independence and that it is meant to be able to apply intelligence at a higher level and speed than that of a human, there is no way they can know for sure what the results will be when they implement it. I think this is a key problem with them trying to move more towards AI based SERPs and the AI simply isnt producing results that we and many users think are best. Maybe one day it will do but it seems it has some distance to go to get to that point.

[StoneSolid]

@Robert

Stonesolid.... I don't think it's so obvious at all.

I know the story isn't simple, and I'm not just talking about de-indexing bug.
I'm talking about extensive amount of illogical things we see from google on a daily basis.

[Mark_A]

My Direct traffic hit an all time high on Monday.
Analytics doesn't seem to offer a way to investigate it any further than the headline.
I don't have the raw logs for that site, but thinking about it I am not certain it would help me.

[ichthyous]

It strikes me that links have somehow become seriously less important after all these updates. I am simply seeing larger corporate sites take over my niches even though I have gotten a lot of new links from very high trust sites recently. It hasn't moved the needle at all and in fact Google is dropping the link count from high trust domains while recording complete spam crap links from scraper sites by the thousands. Even when the scraper site has vanished and I have added it to the disavow list two weeks ago still thousands of spam links reported.

If I do bump up it lasts a day and then I slide back to where I was, but the big downward drops all seem to stick. Nice way to lose 50% of your traffic in six months, despite improving your site and getting lots of new links and five star reviews on Google! My site loads very fast too... But it seems that nothing really moves the needle these days if you are a small business.

[RedBar]

A simple question ... Outside of News sites, is Google actually adding new pages to their index or has it stopped completely?

I added a new section to one of my sites a month ago of 60 pages with all new widget information and images yet only the index page for that section has been indexed by Google and Bing for that matter.

Is anyone else with a regular site(s) seeing similar?

[Zalman]

Indexing has definitely changed over the last year. It's much slower. I try not to force things but after waiting so long I do get impatient.

[Cralamarre]

A simple question ... Outside of News sites, is Google actually adding new pages to their index or has it stopped completely?

My new pages are always indexed by Google. I usually post one or two new articles per week, and I use the URL Inspection Tool in GSC to get them indexed quickly. They're usually indexed and appearing in the SERPs a minute or two later.

I also use the URL Inspection Tool to re-index existing pages when I update them. Google seems to love updated content. Traffic to the updated article often doubles or even triples. Just published an update to an article yesterday and its traffic has already doubled.

[Milchan]

@cralamarre - for your updated content what quantity/percent of the page is new/changed/updated on average? I know there is probably not one size fits all answer, but usually how much do you think it is on the pages that receive a significant increase in traffic? Im interested to know because I imagine there is a sweet spot or at least a certain level you must bypass in order for google to see if as fresh otherwise we could just change a sentence or two and be considered "fresh"

[RedBar]

@Cralamarre

Are you using Wordpress or a custom template?

[Cralamarre]

@Milchan, I usually end up re-writing the entire article from scratch. Most of my older articles were too long, too "wordy", and lacked structure. These days, I'm much better at structuring and streamlining the content in a way that makes it easy for both my visitors and Google to digest. So rather than trying to edit the heck out of the original article, I just write a new, better version. Sounds like more work, but it's actually easier.

[Cralamarre]

@RedBar, I use WordPress. There's not enough time in the day for me to write content full time and be a full-time web developer. So I hung up my web developer hat a couple of years ago and have been relying on WordPress ever since.

[Milchan]

@Cralamarre - ok so in theory it might not be that you get increased traffic due to the content being fresh but more so because the content is actually better - i.e. if you rewrote it completely but did a terrible job of it you could get a decrease rather than increase

[Cralamarre]

@Milchan, I think any time you update existing content, you run the risk of making it worse. But yes, it's not simply a matter of "updating" content, like replacing "Change your hair color in 2018" to "Change your hair color in 2019". You need to find ways to make the original content better at the same time.

Having said that, I usually see an increase in traffic to my articles immediately after re-indexing the updated version. So it does seem like the simple act of updating content triggers something in Google.

[StoneSolid]

Unbelievable amounts of spam, scraped stuff, gateways and such in top results for different keywords in my niche. Really feels like something massively wrong happened with "broad core update".

[ichthyous]

I’ve just discovered (not for the first time) that Google is using some copyrighted custom made images from my site to advertise someone else’s site in featured snippets. And it’s a god-awful site. This is infuriating. Does anyone know if there’s a way to prevent this from happening?

@broccoli make sure you screen shot everything. If your works are registered in DC you may have some sort of case for infringement. After all Google is a business making use of your images for its own purposes. One good infringement suit might cause them to sit up and take notice. Hard to prove though with one incident, would need to be systemic.

[jmorgan]

@SweetPotato

Right, I'm on my own, got it.

I'd recommend reading goodroi's post on "How to Prepare For & Recover From Google Updates":
[webmasterworld.com...]

It pretty much sums up any advice I would give for the most part.

[jmorgan]

Traffic was a touch lower for me in the last day or so compared to last week. I can't find any holidays or events of note. Anyone know why this could be.

I see the spring break is coming up next week, but it shouldn't have affected traffic on a Wednesday the week prior.

[whoa182]

My traffic also seems to be a bit low over the last 2 days. It seems that a lot of non-relevant articles are ranking higher and are from massive brands.

It's kinda funny when you create articles that are exactly 'on point' and directly answering the query, but 4-5 articles rank above you without the keyword in the title and barely any mention of the keyword or are even the complete opposite of the type of product the searcher is looking for. i.e product a without widgets. But they are ranking pages with products that contain the widget.

Google new better how to rank posts before August 2018. They were more relevant.