Google Updates and Who they Target

[martinibuster]

In another discussion someone posted this:

maybe because "global updates" don't exist anymore. It's all about small updates, targeting some of us.

And I will say this:
1. Google Updates very likely do not target spam.

2. If you follow the patents and the latest research, most of what is being worked on focuses on understanding what users want when they type a search query.

3. The SEO community consistently is wrong about Google Updates because they ALWAYS focus on who was targeted.

;)

[Wilburforce]

Is every update to the core algorithm all about catching someone's website?

No, but Panda and Penguin were, and that has probably disproportionately affected the general view. Google DOES target pages that try to game its (information retrieval) algorithm.

I agree that the primary aim is information retrieval, and as the process becomes less dependent on the old SEO stalwarts (backlinks, anchor-text, keywords, keyword-density, word-count) the need to target rogues becomes less of a priority. However, it isn't yet the case that all the old stalwarts are redundant, or that something we once thought of as optimisation won't backfire.

Asserting that Google doesn't target webspam is as misguided as assuming that every update has only that purpose.

[iamlost]

Asserting that Google doesn't target webspam is as misguided as assuming that every update has only that purpose.

Absolutely.
It has always seemed to me that the webspam team(s) are the algo's quality control/assurance backup. Logically, there has to be some human (the algo is ML not even close to AI) oversight. At the very least some calculation/determination of false negatives/positives needs to be made; and is as Google occasionally mentions 'affects n% of sites', etc.

Similarly there is a need for human oversight to notice if there is significant behaviour 'getting around' or bypassing any given desired quality control; this is where - IMO - (1) manual penalties come in as a short term defensive response, and (2) bug reports can be initiated so to eventually remove the need for manual intervention in a given situation.

is every update to the core algorithm all about catching someone's website?

No, but Panda and Penguin were, and that has probably disproportionately affected the general view. Google DOES target pages that try to game its (information retrieval) algorithm.

I disagree somewhat.
Yes, Google does target pages trying to game it's algo BUT it does so either via the algo or via the human quality control aka webspam team mentioned prior. While I don't 'know' I presume that they get their initial target pool primarily from public complaints and some threshold deviance factor of the false positive/negative set.

Panda and Penguin were traumatic to many webdevs, however neither affected a majority of sites; large in absolute numbers and significant in vocal webdev fora members but not in overall query index relative terms. Nor were they about catching someone's specific website. Rather they were simply a raising of n-input thresholds (and probably adding some new ones) such that those running borderline acceptable prior to each black&white update fell below. A straight IR quality control exercise. And frankly, I think they are still set too low. :)

Note: I know that there were false negatives/positives. There always are. That said, in all the publicly shared sites I looked at that claimed to have been unfairly penalised I almost never saw any that were better than borderline. People love their creations; I'm no different.
Note: does Google play favourites? Yup. But rarely if at all with the algo results.

[martinibuster]

Asserting that Google doesn't target webspam is as misguided as assuming that every update has only that purpose.

Not a single person in this discussion is making that assertion.

[aristotle]

In my opinion google's algorithm should give old sites a special rankings boost, because they've proven their staying power and the owner's long-term commitment. Usually a lot of time and work has been expended on these older sites. This is in contrast to the short-lived almost-worthless sites created by churn-and-burn operators.

[anters]

@martinibuster What are you talking about?

You started this thread by declaring "Google Updates very likely do not target spam." and now you post "Not a single person in this discussion is making that assertion." Pick a side and stick with it.

[martinibuster]

Pick a side and stick with it.

Ha! Sorry! Living in New England I've picked up the habit of speaking between the lines.

What I meant was that most of the time updates likely do not target spam. The logic to that statement is that if you look at the research being done today, the majority of it is focused on understanding what users want. The minority of research is on spam fighting. It is the strangest thing because the researchers are proceeding as if the war on spam has already been won.

[ByronM]

If you ask me, Google Search is focused on Mobile space, so most changes in the algorithm are based on mobile experiences - from web results to providing queries and data for its mobile platform and services. (and now for apple Siri)

Beyond that, any changes they do will be around customer knowledge and experiences for the customer.

Oh, and everything they do is on monetizing everything they do. There is no ifs ands or butts about that. To think anything else is silly.

Most of Google's big moves these days are large deals and trades. Google sees huge growth in cloud, so they're leveraging search, marketing and analytics to get people to spend more and trade for cloud services. The product is starting to reflect this "bigger google" in that classic SEO - white or black hat doesn't sustain commercial entities in and of itself anymore, you have to spend and make deals with Google to play on their turf and in exchange, you get more Google data to optimize this relationship.

[Nutterum]

Google does not manually target anything. Hence why all manual updates are highly focused ones. The rest is a big, no, HUGE data results analysis, done with the help of machine learning. These updates aim to provide the general searcher with information on their search with as few clicks to get there as possible. From what I have read through the research papers ( and I want to thank martinibuster and other industry analysts here!) all updates and the majority of patents have this one goal in mind. To make search more ordered, classified and interconnected. Hence why through time some brands saw major decline in organic traffic, simply because Google's Rank brain and other machine learning algorithms, determined that this brand should not funnel traffic from content spaces not linked properly to the space these companies operate in.

It is the very same reason algorithms like Facebook, who can't really use such techniques have harder and harder time to distinguish between proper content and content building to funnel impressions and clicks from unrelated niches.

It is this same reason why, specialized social networks like Pinterest, often ranks higher on Google for products, fashion niches. Why Amazon rank best for product resellers and why e-bay ranks best of auctions, or location review agregators rank best for food/service companies. All these online websites show everything the user WANTS straight up, no clicks, no "see more" buttons, no pay walls. Just reputable content.

Again, I am not discussing whether the system is perfect as it is not for manufacturers. I have battled this battle more than once to know how hard it is to rank for "widget x" even though I am the main manufacturer for the geo-location I am operating in.

All I am saying is that in my mind all discussions that do not contribute to the "easy content now matra" that Google and their services aim to be best at should be taken with a pile of salt.

[martinibuster]

Hence why all manual updates are highly focused ones.

Great observation!

Manual updates seem to have a surgical quality to them, in that they not only want those sites to not rank, but they want them completely out of the index.

[System]

The following messages were cut out to new thread by goodroi.

New thread at: Using the age of a site as a ranking factor [webmasterworld.com]

[edited by: goodroi at 9:42 pm (utc) on Oct 24, 2017]

[wadsworth]

OK, a few ideas I'd like to contribute:

1. Take care with on-page monetization features (not only ads)... read your QRG
2. Check your internal linking on your site, are you linking to your best pages with useful, relevant, quality links?
3. User experience elements, user interface, navigation elements, how easy is it to navigate and understand your site?
4. Are you providing relevant, quality supplemental content?
5. Log analysis - do you know what Google is crawling on your site? CMS can create spider traps, watch out for dynamic pages that generate more dynamic links, watch out for your search pages creating more and more search pages with dynamic links

If you've got everything on point technically, and you lost big rankings, look at who beat you, what are they doing better?

[FranticFish]

Your opinion would have more credibility if you could dig up patents and research papers to show that Google's researching algorithms to increase ad clicks from their SERPs.

As if these would ever be published if they existed. Please. I'm NOT saying that they DO exist, but I am saying that your dismissal - out of hand - of the possibility of some internal tampering shows a level of trust I do not detect in your analysis otherwise.

Shaddows made the point that SERPs design (business practice) is separate from the algorithm. Is it so hard to believe that they would apply some other - undisclosed - factor to the results to suit business purposes? Of course it would be the end of them if it were publicly known. But that's not the point.

It has been oft observed that GSC/GWT and Analytics ranking data doesn't agree. I've seen rankings reported that did not match real world rankings. Why? In that case, it's been posited that results come in 'raw' and are then filtered to apply other factors - such as penalties. This indicates that results are processed before being exported to the SERP. So the mechanism exists to tamper with the 'raw' SERPs.

Are you honestly saying that you know - for a fact - that Google doesn't tamper with the results for the biggest money terms to increase ad revenue based on the absence of patents?

I do realise that this is not a road to go down because blame (right or wrong) isn't productive and doesn't lead to a solution - and I get that that's what you're focused on.

But I don't think that those that suggest tampering should be ridiculed or dismissed as conspiracists or nutjobs. You may remember that VW were caught interfering with their engines to comply with emissions tests. Did they patent that?

[martinibuster]

CLARIFICATION

My original post did not convey what I thought I was communicating.

This

1. Google Updates very likely do not target spam.

Was meant to express this:

1. Google Updates very likely do not target spam with every update.

In other words, some Google updates could be changes in the way an algorithm calculates what is relevant to a user when they make a query, resulting in some sites dropping in the SERPs as a result. Sorry for the confusion.

You may remember that VW were caught interfering with their engines to comply with emissions tests. Did they patent that?

As a matter of fact, VW did patent the formula [mriplaw.com] for complying with emissions tests.

I'm not dismissing out of hand the possibility that there is an algorithm to balance out user satisfaction while maximizing ad clicks. I am only saying that the hypothesis would have more credibility if there were research done on that kind of algorithm.

It is entirely possible that proof of such an algorithm exists and no one has noticed it yet. I would love to find it and have in fact searched for it but so far have had no luck. I truly would like to find it if it exists. It would be huge news.

Afaik, such research does not exist to support that theory. If a web publisher is to spend their time making business decisions about their Internet Marketing, it is only reasonable that they demand that the theories they base their decisions on have at the very least a known possibility of being real.

Otherwise you are dealing with a baseless opinion.

The research or patent doesn't even have to be by Google. It could be by a university researcher anywhere on the planet or Yahoo or Microsoft. In order to raise a theory from pure opinion with zero foundation to a credible theory, you must have some kind of proof that there is a possibility of it actually existing.

I do not claim that such and such theory is in actual use. I'm only saying that such and such is possible when a prior patent or research paper exists. Failing that, a theory that lacks even a hint of a foundation, is pure speculation.

[goodroi]

@martinibuster It looks like you made a mistake and did not fully read the articles that you reference. The VW patent in that article is less about complying with emissions tests and more about providing a cover-up for their actions. Here is a quote from the article

For example, prosecutors may argue that VW was attempting to justify their reasons for reducing reductant injection at lower temperatures (off test cycle) as a form of "protecting the vehicle against damage."

It is not wise for any SEO to quickly dismiss things unless they are provided overwhelming evidence. Lance Armstrong kept saying he was clean until he was finally proven to be doping. Yes, there are often patents that can be linked to SEO changes. But this is not always the case. Some patents are quite confusing and hard to link to specific business actions. Other business actions are never patented but happen everyday. Smart ideas can't be patented but they sure can make big money.

It is smarter to keep an open mind and realize that all of us can learn new things. The goal posts of SEO keep moving and we need to stay flexible and adapt if we want to survive & thrive.

PS If anyone wants to research what other smart people are researching head over to [scholar.google.com...] it lets you more easily discover new research papers & patents for any industry.

[martinibuster]

The VW patent in that article is less about complying with emissions tests and more about providing a cover-up for their actions.

You know... The point of that link is that there was a patent. Their motivation for filing it is irrelevant to that point. ;)

It is not wise for any SEO to quickly dismiss things unless they are provided overwhelming evidence.

Slow down there, partner. I have already stated this in very clear words: I am not dismissing the notion.

Can you hear me now? ;)

The point is, and it still stands, that as far as we know at this time, there is no credible basis for the claim. No foundation. I'm not dismissing the notion. I'm just saying that at this time, it's just a notion, an opinion of what's going on at Google until some evidence can be found to elevate that opinion to a possibility.

It could very well be that the patent that could provide credibility for the currently baseless theory that Organic SERPs are engineered to provide more clicks to ads is hidden within a patent for optimizing clicks for AdWords auctions. But as far as we know, the evidence does not exist. As I stated already, I would love to find it if it does exist.

[goodroi]

@martinibuster The point is that you incorrectly represented the VW patent while trying to refute another post. Facts matter. When we accidentally mislead or intentionally misstate facts it does not help. We should all welcome clarification and corrections as we work together to better understand this crazy SEO world because we all make mistakes.

You have stated "You are not dismissing the notion" but you have also made comments like "Otherwise your theory is easily dismissed." So your position might not be clear as you may think.

You also said "there is no credible basis for the claim". That is not accurate. You can say someone has not yet linked to a patent to your liking but to say it is not credible to believe a for profit company would possible try to profit is an inaccurate statement. Yes, some conspiracy followers may take that idea to an extreme & inaccurate place . But on the flip side we also need to be careful we don't become extremely naive & blind to Google's for profit business. As often in life, reality & truth generally lies somewhere in the middle of the extremes.

Everyone: Let's try to keep an open mind and respect differing opinions which might not match our own personal opinions :)

[martinibuster]

...but to say it is not credible to believe a for profit company would possible try to profit is an inaccurate statement.

Again, you're misrepresenting what I stated.

I didn't say the theory was not credible. I stated there was no credible basis.

I did not say I was dismissing the theory, I stated that without a credible basis to support that theory that it can be set aside as just an opinion, hence, easily dismissed. <snip>

[edited by: goodroi at 3:23 pm (utc) on Oct 27, 2017]
[edit reason] Let's stay on topic & Keep WebmasterWorld Community Center posts in Community Center [/edit]

[EditorialGuy]

Let's also remember the difference between a hypothesis and a theory. (A hypothesis is conjecture, while a theory has been tested rigorously.)

[seoskunk]

What utter nonsense about patents. Google are hardly going to patent illegal non competitive behaviour. I found these quotes online regarding the recent case in the EU against Google. This is neither hypothesis or theory, its people commenting on the facts.

What Google has done is illegal under EU antitrust rules. It denied other companies the chance to compete on the merits and to innovate. And most importantly, it denied European consumers a genuine choice of services and the full benefits of innovation.

-Margrethe Vestager

The prohibition of Google’s immensely harmful search manipulation practices is far more important. There can’t have been many competition cases where the stakes for consumers, businesses, and innovation were any higher.

For well over a decade, Google’s search engine has played a decisive role in determining what most of us read, use, and purchase online. Left unchecked, there are few limits to this gatekeeper power. Google can deploy its insidious search manipulation practices to commandeer the lion’s share of traffic and revenues in virtually any online sector of its choosing, quietly crushing competition, innovation, and consumer choice in the process.

-Foundem

Part of the EU vs Google case involved the Panda Update, this algo specifically targeted Google's competition on comparison shopping. Between 2007 and 2013 nine rivals to Google Shopping lost 79% of their traffic, meanwhile Google's own shopping service increased traffic by over 200%.

Now there is your credible basis.


[techcrunch.com...]

[europa.eu...]

[theregister.co.uk...]

[martinibuster]

Good post seoskunk! Excellent! :)

Let's discuss that. ;)

Google's response: [blog.google]

While some comparison shopping sites naturally want Google to show them more prominently, our data shows that people usually prefer links that take them directly to the products they want, not to websites where they have to repeat their searches.

That's similar to why Directories stopped ranking years ago. Remember when that happened?

Google's explanation for this demotion is the same as the decision to demote directories (as I recall). The reasoning was that Google said users prefer results that take them straight to the answer they're looking for. Users do not prefer a link to a site where they have to repeat their search. Even DMOZ was demoted in the SERPs.

So is that an example of targeting competition through their algorithm in order to gain more exposure for their advertising?

Or is it Google targeting sites that do not directly answer a users query?

[seoskunk]

While some comparison shopping sites naturally want Google to show them more prominently, our data shows that people usually prefer links that take them directly to the products they want, not to websites where they have to repeat their searches.

I don't want to stray too far from the point of what Google Updates target, but Google response seems inconsistent when their results show plenty of examples of content such as image and financial comparison sites that do not take users directly to the original content.

That's similar to why Directories stopped ranking years ago. Remember when that happened?

Part of the EU Investigation into Google centres on localised search which is normally driven via directories. Google launched its own directory service for local business (initially driven by data from 'rival' directories) and then produced a series of updates that effectively removed directories from its search.

So is that an example of targeting competition through their algorithm in order to gain more exposure for their advertising?

Good question, we await results of the EU Investigation. One directory service 'Yelp' have made complaints to the EU and FTC according to Wired regarding how Google populates its own directory data.

Yelp says it found Google pulled almost 386,000 images from Yelp in an hour, and then used some of the photos in business listings in Google Maps. Yelp says it searched Google for 150 of those businesses and found that a Yelp photo was a lead image in Google’s Local OneBox—which shows a business’s location, phone number, and reviews—in 111 cases.

[wired.com...]