Google Updates and SERP Changes - August 2017

[MrSavage]

---

System: The following 12 messages were cut out of thread at: https://www.webmasterworld.com/google/4856114.htm [webmasterworld.com] by goodroi - 8:21 am on Aug 2, 2017 (utc -5)

---

Google is, or should I say has, just become or is closely becoming an actual website. You know. The place you go to get what you need and not go somewhere else. It's almost officially the era of the Google tax. They got a piece of everything that comes through the web.

[aristotle]

Looking at the logs for one of my sites, more traffic seems to be coming through googleweblight lately. It comes through google proxies, so I can't tell where its originates. Might be spam except that it behaves like real humans.

[Mark_A]

Thing is, there can only be 10 pages listed in the top 10 SERPs. That, by definition, means that perhaps 100s or 1,000s of webmasters are going to be grumpy that their page isn't there.

My experiences of users, including many who use adblocking, is that they are nowhere near dissatisfied with Google search results in fact they are quite happy - so there is no likelihood at the moment of duckduckgo coming to the rescue of webmasters by delivering the sorts of volumes of visitors which Google can deliver.

[Shaddows]

they are nowhere near dissatisfied with Google search results in fact they are quite happy

Absolutely. This is the thing with AI- it will give users what they want (their revealed preferences, in the jargon), rather than what's good for them, or what is critically better.

So, you may be perfectly correct that your site is better than the thin, superficial, ad-stuffed MegaCorp site, but if users click on the crap regardless, Google will return the crap.

[samwest]

Since I surprisingly hit my weekly conversion quota 2 days early, I'll wager that I do not see another conversion for the next two days. Place your bets. These quota cases make great evidence of how the system is now working, doling out results in exact measure.

[browndog]

So why did Google push for webmasters to create excellent content if they're going to ignore it and return results for mega/global sites with thin content?

I remember Matt Cutts (and I'm not having a go at him when I say this) telling us to create great content. To imagine it's going to be published in a magazine (I think that's what he said). Every time I write an article, I have his words in my mind. Imagine it will be published, it must be absolutely spot on, include lots of images to help the reader identify with the topic. I always shoot for that, because Google told us to do it, but return tripe in the results.

I think many users probably are happy, but maybe they don't know there's much better content out there. If I was searching for information on how to change a tyre on a motorbike (I imagine motorbikes are the topic I know least about), I woudln't know a good result from a bad one. But, the good one might give that extra bit of info like the best tyres, if you do X you may damage Y etc., but the thin article might not. As a novice, I'd be none the wiser.

[sikosaurus]

So why did Google push for webmasters to create excellent content if they're going to ignore it and return results for mega/global sites with thin content?

That really is the question of the day.

They should rename and rewrite "Google quality guidelines" to "Guide for building undetectable private blog networks"

It would certainly help new webmasters to start on the right path asap.

[EditorialGuy]

So why did Google push for webmasters to create excellent content if they're going to ignore it and return results for mega/global sites with thin content?

In my experience, Google isn't ignoring excellent content, and "mega/global sites with thin sites" (a.k.a. content farms) aren't doing nearly as well as they did a few years ago. As always, YMMV.

[masterjoe]

Surprise Surprise.

Samwest, I had a long string of conversions over a few hours and just met my quota too. Now its back to useless traffic yet again.

[browndog]

In my experience, Google isn't ignoring excellent content, and "mega/global sites with thin sites" (a.k.a. content farms) aren't doing nearly as well as they did a few years ago. As always, YMMV.

True, results may vary depending on the niche. My main competitors aren't content farms, one is a large media company, and their content isn't bad, it gets a bit annoying when they get multiple results for the same search topic (up to four on the first page of Google), the main site I'm referring to is an international food company who write thin content about the widgets they cater to, not even food related, just widget care in general. 120 word articles which offer very little.

Interestingly, when I read your response I was on my phone, on the train, and my search is still Google on the phone. I had to look up YMMV as I'd never heard the phrase, the first result was a site so full of advertisements that I couldn't find out what YMMV means and hit the back button. So not great results there when you can't actually find what you are looking for due to a page jam packed full of ads. But Google thinks that page is worthy of the no. 1 position.

[glitterball]

What I see nowadays is that Google returns results from major broad-topic websites above specialist websites, because it is easier for them to guarantee a "good enough" result for the user.

If Google tried to identify excellent expert websites, there is a much greater chance that very low quality pages would slip through. So from G's point-of-view there is less risk associated with returning mediocre results.

I didn't believe in quotas before, however the charts in Google Search Console are way too consistent for there not to be quotas.

[samwest]

As predicted, not a single conversion over the past 48 hours once weekly quota was hit. SERPS haven't changed one bit, only the apparent quality of traffic. Week to week variation in conversions remains at +/- 0.5 or less. Traffic always starts out varied, but zeros in to nearly identical totals by week end, with slow starts playing catch up and fast starts being hobbled. It's just like a PID control loop. The level of perfection suggests the presence of a generative adversarial network (similar to StackGAN) in operation, except this unsupervised learning network appears to be trained to optimize traffic and conversions, not images.
It seems to work so good that it can "debounce" any trends of weekly surges in traffic, real or forced.

[masterjoe]

Yep, seeing the exact same pattern. This is ridiculous. I suppose we should have all seen this coming though. Now that Google has the power to change everything and is slowly but surely becoming a "one stop" destination (monopoly), they can now freely screw over anybody with almost zero repercussions. I'm glad the EU gave them a huge fine, although that won't matter once they are your bank, communication service, tax agent, the robots at your grocery store, whatever. With the amount of information they steal about everyone they possibly can, it is mind blowing how far this manipulation is going.

The monopoly needs to be stopped. I have been on YouTube watching some unrelated videos for once, it is difficult to find good content that you want to see. I've also noticed a lot of people I was subscribed to a while back (haven't visited in a long time), that a lot of my creators are using something called "Patreon" -- a service which allows people to pay a one time or subscription fee. Why? Because they screwed over the people who have spent hours upon hours creating content and paying them significantly less than they used to. Its pretty clear where this is all going. Get access to people, get a majority of the marketplace, screw the people who made you what you are.

Much like Amazons culling of commission increases for affiliates, Ebay's bs fees, PayPals rip off conversion and transfer rates. All these companies are making the internet a sucky place to be.

[browndog]

If Google tried to identify excellent expert websites, there is a much greater chance that very low quality pages would slip through. So from G's point-of-view there is less risk associated with returning mediocre results.

I liken it to McDonalds. You know the quality of the food will be the same, no matter where you are and it satisfies a need (hunger), but...there are much higher quality but lesser known restaurants out there which will provide a considerably better experience.

[SnowMan68]

I liken it to McDonalds. You know the quality of the food will be the same, no matter where you are and it satisfies a need (hunger), but...there are much higher quality but lesser known restaurants out there which will provide a considerably better experience.

This is a great example.

Many people will choose to go to McDonalds because of its brand name recognition (similar to people clicking on large branded sites being shown in the SERPS). This is especially true when people are in different cities that they don't know very well (new verticals they are searching in). Your restaurant needs to be AMAZING to standout to new users, because your brand has zero value to them. Similar to how Google likely views it. When you leave McDonalds, you're probably fully and satisfied. Which means you also spent some time there and Google recognized this as a good experience (even if it was just okay). For most people, when they walk through the door at McDonalds, they are already know what to expect. With small unknown establishments, it's a dice roll. It has the potential to be dirty, maybe the service is slow, maybe it doesn't feel right. Either way....it's unknown until you spend some time there. The thing with the internet, it's a lot easier to go to a different website vs. bouncing around the city to find a better restaurant. If you are not making the right impression immediately and people are leaving right away....you are sending a horrible signal to Google. Because most people who walk through the front door of McDonalds will eat there.

[glakes]

Thing is, there can only be 10 pages listed in the top 10 SERPs. That, by definition, means that perhaps 100s or 1,000s of webmasters are going to be grumpy that their page isn't there.

The problem occurs when those ten positions are occupied by 4 or 5 different organizations. When this happens, the core reason why people use a search engine is lost. And that reason is being able to find sites on any given topic that one would otherwise not know about.

I liken it to McDonalds. You know the quality of the food will be the same, no matter where you are and it satisfies a need (hunger), but...there are much higher quality but lesser known restaurants out there which will provide a considerably better experience.

Very true. And the chances of Google producing the wrong search results, or omitting what you ordered in the query, are about the same as going through a McDonald's drive thru window and receiving missing items from your order. At least Google is a bit healthier.

[Jori]

What's the purpose of Google, if it keeps sending me results I would have found by myself? I don't need help to find Amazon, thank you!

What's the purpose of sending me 3 or 4 links to the same company? I thought that 3 or 4 pages with similar content in the same website would be penalized by Panda!

I do see huge forums, with no credibility in "money or your life" subjects, in front of a lot of old and well established websites.
Why? Because they speak about the same subject in 3, 4 or more different topics. It's like this for medical subjects, loan subjects and other very touchy things. Where is the quality guidelines in this?

Is it normal? How this kind of results is handled by AI and machine learning?

[EditorialGuy]

What's the purpose of Google, if it keeps sending me results I would have found by myself? I don't need help to find Amazon, thank you!

On the other hand, Google would look stupid if it didn't include Amazon pages in its results. And sometimes those pages are really useful even if you aren't shopping. If you're looking for information about an obscure product, Amazon may be the only useful review source out there. Sometimes Amazon's pages are even helpful for tech support: I've often been amazed by the kindness of strangers on Amazon. (And I'd be annoyed if such pages weren't included in Google's search results.}

In any case, most searches don't involve e-commerce, and most queries aren't likely to be satisfied by searching or browsing on Amazon. If you're looking for a biography of St. Catherine of Siena, information about the moons of Jupiter, or advice on riding public water buses in Hamburg, Google is a great place to begin your search.

[RedBar]

In my widget sector the current results are the very worst I have ever seen to the extent that they are pointless.

A few months back I reported here about a website that had been ranking #1 for many, many years that simply disappeared but then reappeared a couple of months later and is still there.

Well, G's done it again to me with another site, keyword1keyword2 dot com. It's been at #1 for years yet now when typing in the keywords, it can't be found anywhere in the SERPs. Bing still has it at #1.

It's indexed so what have they done with it?

Honestly, being polite, this lot couldn't organise a party in a brewery!

[Jori]

EditorialGuy : and 4 results from Amazon about a generic product I didn't asked for? When I type "blue widget" I really mean BLUE, not green.
If some obscure little website talks about blue widgets, why on earth G is keeping showing me 4 results of green widgets from Amazon?

I'm getting paranoid.

[EditorialGuy]

When I type "blue widget" I really mean BLUE, not green.

When I get off-topic results, it's usually because one word in a multi-word query is off topic. To a human, that word may seem critical, but to an algorithm, it may just one more word in half a dozen--at least in the current state of the art.

To use a purely imaginary example, if I searched on "how do I cook a cheddar and bacon omelette," I might get a result about cooking a cheddar omelette or a bacon omelette, but not necessarily a cheddar and bacon omelette. As someone who likes to eat, I see a big difference between an omelette with only cheddar or bacon and an omelette with both, but an algorithm or AI black box may not grasp the importance of "and" in my query.

Unfortunately, a search engine is always a work in progress, so I've learned to control my expectations.

[Jori]

Yes, you're right. But, and that's new, it seems to me that not so long ago, Google would return me my blue widget result. And your omelette with cheddar and bacon. Because it would give me small websites instead of big players in the widget industry or cook industry.

This is what I'm seeing, but I'm trying to understand what the algo likes. We are like 15 years back, and cheating Google is stronger than never now. Maybe it's just a question of time, but if the algo can not see the difference between an omelette with cheddar and bacon and an omelette without one of those two ingredients, what's the meaning of this supposed artificial "intelligence" ? Just give me a page containing the keywords I asked for like in the good old days!

[seoskunk]

'how do I cook a cheddar and bacon omelette' - i get good results for this tbh

[browndog]

Many people will choose to go to McDonalds because of its brand name recognition (similar to people clicking on large branded sites being shown in the SERPS). This is especially true when people are in different cities that they don't know very well (new verticals they are searching in).

Yes, if I was on the road and needed a quick bite, I'd trust McDonalds over a diner (I have actually had food poisoning after eating an egg sandwich from a service station type cafe so my experience may be somewhat skewed). I don't particularly like their food, but it serves a function and I do trust the brand.

But, if I want a nice dinner for my birthday, McDonalds wouldn't even get a look in. But it seems because it's so trusted, Google return McDonalds type results and don't include your more quality restaurants (sites) which provide a much better experience. They think McDonalds serves its purpose and it possibly does for 85% of searchers, but they should be given some choice.

Over and above the big brands with thin content, I am really sick of seeing domain clustering.

[aristotle]

I might get a result about cooking a cheddar omelette or a bacon omelette, but not necessarily a cheddar and bacon omelette

Most likely they would put the cheddar omelette in the answer box

[samwest]

'how do I cook a cheddar and bacon omelette' - i get good results for this tbh

Yeah, but I never asked for ONIONS! (but that's what's in the answer box recipe....)
;)

[Awarn]

I have recently seen an odd change I have been focusing on more detailed descriptions of 11K products. Under most of the sites that carry these products the products are sorted under Method A or Method B. It has appeared that Google was picking pages based on Method A. Most users will find more specific results if they follow method B. So in many descriptions we have been putting in a link to see similar items that show up based on method B. Now Method B uses a distinct word. I was bouncing from the top listing on page 2 to bottom 2 listings on page one when it appeared the listings were based on method A. No mention of Method B listings anywhere.

So I add a lot of content and show this method B that uses a distinct word expecting gradually rise. No what happens is the distinct word shows up in the listings and I actually drop 3 slots. All page one listings switch to show method B listings. Almost as if Google is learning from my site and using the information to rank others but doesn't evaluate me in the same light.

I am on some strange quota from the data on WMT. The sitemap exists and has for years. It never shows like indexed amount is correct. About every 10 days the indexed amount increases almost exactly 300 pages. This is messed up.

[Jhurwith]

Anyone else see a minor update on Friday (18th) or Saturday (19th) of last week? I am trying to determine if we were hit by algo or seasonality?
Thank you.

[SnowMan68]

Yes, we saw a bump of about 15% on the 19th.

[renatovieira]

A huge drop in my traffic since last week. And it keeps falling. So far the fall of 80%.

Someone else?

[Jhurwith]

Anyone know what the changes focused on?