Exploiting recent Google ranking mechanism

[john1444]

Don't know if your findings hold any water but it looks promising and real smart. Whatever it is, you need to be on the first page of Google using the conventional SEO practice before algo detects the peoples choice..

 

[-0o0-]

Ok thanks, Ive set up a few domains for testing. Now, please move this to JNR VIP section

 

[abyssal_miss]

does this work without emd's? you said buy a new emd. What about existing non emd's?

 

[mynameis940]

Can you post some example links here? Just put the URL and TLD separately.

 

[pxoxrxn]

Anyone have any updates on this?

 

[hnmfvnxe]

Ok thanks, Ive set up a few domains for testing. Now, please move this to JNR VIP section

Yes hurry please....because google cannot afford $97...LOL

 

[YouFeelMeDawg?]

Sorry to disagree, but your whole theory on this is wrong. The short verse long click is not as simple as your put it. The math equation that Google would be using would be more like this....Some % of clickbacks to Search, for phase AAA, on a site of at least XXX size, with ZZZ amount of content on page [ content sub variables B, C] , and on a page level, the page having properties of Q, W, l in some combination. Now, there is a really smart guy working over at Google named, Biswanath Panda, the name might sound familiar to some here as the Panda Update was in fact named after him. This guy wrote a very interesting academic paper called....PLANET: Massively Parallel Learning Of Tree Ensembles With MapReduce. Suggest you have a read, a lot of good information there about classification and regression tree learning on massive datasets.

I completely agree with you 100% on this.They also apply statistical models for their anti-spam algorithm on youtube comments, especially for determining a thumbs up or down being valid or invalid.

Google algorithms are heavily centralized around "Ensemble Trees".To be honest, I would suggest anyone that is serious into using the right scientific way to break their algorithm or atleast come close to it, to actually collect a bunch of data and use statistical data compressed models.



The paper your describing is pretty much using clusters of computers put together,sort of like Hadoop but using their own , MapReduce

To the rest believing that fairly tale shit, get real.

Google has mathmaticians using statistics and state
of the art tree learning algorithms.

By the way, that paper is pretty much one of the most famous papers from a google engineer. It is pretty old, but its extremely popular even among non-academic people.

 

[grahamklemm]

Thankyou, glad someone read my post on this.

 

[mickyfu]

So if you do not get anyone to click on the first site, the second will not rank. I see no point in this or am I reading it wrong. You have to rank the first site for people to click through to the second site. Why not just rank the second site. I read what you put about EMD's but in my opinion it is completely wrong. EMD's no longer get ranked high, I use many EMD's a month and have never just seen them on the top spot on weight of their EMD. Plus another major flaw is the fact no matter what niche you go for, all the emd's are taken.

 

[grahamklemm]

In fact, it seems what you are talking about here is creating some sort of doorway page. If so, both your sites will probably end up getting banned. Do a search on Google and have a read on what happened to GermanysBMW when they tried this.

 

[mickyfu]

Have you checked if these gateway pages are being blasted with links and the gateway is not just there for a backlink. Because the way I see it, what you have written makes very little sense. All sites have links to them and people follow these links to sites. No difference in making a gateway page. In fact if what you say works, you would be best starting a thread on a forum linking back to your site and get click throughs this way.

 

[phpbuilt]

Another thing, exact match Domains have ZERO effect on rankings in Google. From John Wu at Google...."Exact Match domains does not equal rankings for that keyword phrase".

Oh come on, you got to be kidding me.

First off, logic fail. If someone goes and buys "casinos.com", does it automatically equal rankings for casinos? No. It still needs content, backlinks, etc. to rank (hence does not equal rankings). Does it have a bonus towards ranking for that term, assuming it isn't a low quality site? Yes it does.

Second logic fail, assuming any employee at Google is going to give you the full, complete, not-subtly-confusing, iron-clad truth about ranking. Not going to happen, ever. Sometimes they tell the truth, sometimes they lie, sometimes they tell partial truths, anything a Google employee has to say about ranking should be taken with a grain of salt.

As long as we're quoting Google employees and standing on it as unabridged truth ...

https://twitter.com/mattcutts/status/251784203597910016

Minor weather report: small upcoming Google algo change will reduce low-quality "exact-match" domains in search results.

Matt Cutts both implies that exact-match domains have had a bonus, and the bonus has been reduced ... but only for "low quality" domains. Also implied is that "high quality" domains remain unaffected, thus keeping whatever bonus they previously had.

I, on the other hand, believe Google employees lie as a method of preventing the SERPs from being manipulated.

http://searchengineland.com/live-blog-matt-cutts-smx-advanced-43906

Bounce rate and rankings? Matt says Google Analytics is not used in the general ranking algorithm. ?To the best of my knowledge, the rankings team does not use bounce rate in any way.? He tiptoed around this question a bit, choosing his words very carefully.

Why wouldn't they? It really is an awesome statistic for determining whether a page has quality. If a page has 90% bounce rate, and another has 70%, why wouldn't Google use that as a ranking statistic?

Of course they use it. But they don't want you to know that. Why? Because the next logical step is that if bounce rate has an effect, it can be manipulated. Google doesn't want it manipulated ... in fact, imagine the chaos that would ensue. If it were a proven fact that bounce rate had an effect on ranking, blackhats would start scripting searches on Google, consuming resources ... costing them money and having results manipulated (which they don't like). So, they lie about it.

Now here comes a guy forward and he doesn't have theories and guesses. He's got evidence. Working examples. An exact technique. I'll take that over pure guessing any day (and yeah, for as much as you name dropped Google employees and referenced white papers, you're still guessing).

 

[YouFeelMeDawg?]

Now here comes a guy forward and he doesn't have theories and guesses. He's got evidence. Working examples. An exact technique. I'll take that over pure guessing any day (and yeah, for as much as you name dropped Google employees and referenced white papers, you're still guessing).

The OP hasn't posted any proof at all.
May I know what proof your talking about, since I absolutely don't even see any proof from OP where this is taking place. In reality, the OP is making it seem like the CTR is the only reason why these sites were ranking.

(and yeah, for as much as you name dropped Google employees and referenced white papers, you're still guessing).

Lol.... your going to take over CTR non-proven hack over white papers on state of the art tree learning algorithms.

 

[Berkeli]

This isn't gonna work as simple as that. but yes Google does track search experiences but it is no news to anyone.
reasons why this method is just not gonna work:

1. Google won't care about clickthroughs, e.g. my business (in addition to billions of other pages) doesn't have clickthroughs
2. Google tracks time on page, returning visits, bounce rates, sites loading times, server location (NOW THESE REALLY AFFECT YOUR SERPs)
3. If your visitors spend 10 secs on your page and then get redirected/click to another site, your page is never gonna rank on google. same as bit.ly/****** links. WHY THEY DON'T GET RANKED on google ?
4. Without linkbuilding you are never going to rank either.. I have bit.ly URLS ranking on first page of google with good link building ( were used in one of my campaigns on TIER 1 as masks for money site, results are EPIC!)

 

[grahamklemm]

I disagree with everything you saying.
First, read what Matt Cutts and Google are saying and look at all the available evidence. At NO stage has anyone from Google said there was a bonus is an exact match domain in the last few years. What they are clearly stating is that Exact Match Domains do NOT equal ranking for that keyword phrase. Now, the Twit from Matt states that they will reduce low quality exact match domain in the search results...hence, they are applying some sort of penalty aimed at those who attempt to game the system using exact match domains on poor quality website which is exactly what this user has posted. If you read his post he is talking about using an exact match domain on a single page with little content that is to be used as a doorway page. Now this is the exact sort of thing that Google keeps warning people about and the exact sort of thing that will cop a penalty or two or at the very least a massive downgrading in the serps to the neverland. Also, you state he posted working examples...Where? I see no case studies here. I see no evidence here of his theory. Its a theory and its a bad one.

Now, as far as what you said about bounce rate your wrong. Bounce rate on its "own" could not possible have any effect on the serps. But if you take that as part of....some % of clickbacks to search for phase AAA, on a site of at least XXX size, with ZZZ amount of content on page content sub variables B,C , and on a page having properties of Q,W,L in some combination then what you will have is a TRIGGER for google to take a closer look at your website to decide if your site should be penalized or banned.

 

[ija61]

First of all please describe what tests bring you to this conclusion.... How many websites did you manage to rank with this....

1)
Now beside it have been discussed all over I also have notice the EMD doesn't work as great as they did.... So if you want to rank an website for "online seo" if you use a EMD will not guarantee that you will be on the first page.

So you still must rank to get you EMD on first pages, and doing that for a competitive kw is pointless since it will be better to rank your website instead of the EMD.

Now if you try to say that ranking an EMD for "online marketing providers in Guatemala that speak English" and place a link with "online marketing" to a second website will boost the last one on first page for "online marketing" I think you are joking

2)
Let's say I'm using your theory and I use a EMD to rank for "online SEO" and on that I will build a link to a second webpage with "SEO" as keyword since I try to rank my website (the actual website) for "SEO"......

I don't think G will give me more authority due to those clicks that came from a link with "SEO" as anchor text, more than a website that are already listed and displayed on searchs and get click from direct traffic that search directly for SEO.

3)
How about anchor text variation....? This will leave a BIG footprint... And this if we not take in consideration that we are talking about visitors that are login in their G account so G have a great visibility.

4)
I don't think the time spend on site have a great impact because this is very different depending on the site. On a blog you will stay to read the content but on a file sharing site you will not spend so much time.
Second the time a visitor spend on a forum is much higher than the time the same visitor spend on a blog.
The same happen with eCommerce websites....

5)
Why should I try to rank a second website and redirect visitor to mine when I can rank my own...
Because if you want that a visitor stay on your website after he click the link on the first website the second one must be related to the kw he have first searched for, if not he will get out in second...
And if the kw is related to my main website why don't try to rank for that?

 

[phpbuilt]

The OP hasn't posted any proof at all.
May I know what proof your talking about, since I absolutely don't even see any proof from OP where this is taking place. In reality, the OP is making it seem like the CTR is the only reason why these sites were ranking.

Go read the OP again. You'll clearly see that he said message him for domains which are ranking using the method.

Lol.... your going to take over CTR non-proven hack over white papers on state of the art tree learning algorithms.

Debate fail. Do you see me anywhere confirming what the OP posted as being accurate? No. Interesting theory to check out, yes. So I'm not "taking CTR non-proven hacks" over anything, proving it is the point of the thread.

In fact (as you said), OP only believes CTR and EMDs to be an influence. But instead of talking about CTR and EMDs, the thread got hijacked and alternative theories on why things rank were introduced by grahamklemm and you.

Then you took the next annoying step of assuming that anyone who believes CTR and bounce-backs may have an effect (as OP states), or that EMDs have an influence (as OP states), ... that those people or anyone who would consider such things as being inherently against the white papers, patents and name-dropped Google employee quotes that you introduced. Quite frankly, they have nothing to do with each other. There's a multitude of considerations Google makes in ranking a website, and you can both believe in EMDs having an effect, and CTR/bounce-back rates having an effect, AND every other little thing you guys mentioned as having an effect, without anything contradicting each other.

Most of it had nothing to do with CTR or EMDs. What little that you guys said that actually applies to EMDs (such as the John Wu quote) ... are clearly inaccurate and taken out of context. (For instance, saying John Wu says EMDs have zero effect when, as quoted, said nothing of the sort).

Honestly your guy's posts are nothing short of a thread hijack, dismissing all claims of the OP and injecting your own alternative theories on why things rank.

That being said, I haven't contacted the OP. I do, however, believe EMDs have an effect on rankings (even now after the Google EMD update), and I do believe CTR/bounce rates have an effect on rankings (http://forums.seochat.com/google-optimization-7/ctr-bounce-time-site-google-search-engine-optimization-426500.html http://www.seomoz.org/q/are-click-through-rates-bounce-rates-as-ranking-metrics http://www.seomoz.org/blog/the-2-user-metrics-that-matter-for-seo). Its not a far jump to consider what the OP says has merit. Its worthy of testing for those who have the time/means.

 

[phpbuilt]

At NO stage has anyone from Google said there was a bonus is an exact match domain in the last few years.

TL;DR

But I will say this from glancing briefly.

Here you are again above, trying to look for a quote from someone at Google saying there was a bonus for having an exact match domain in the last few years.

Really, WTF? Haven't you even tried? Why do you even need to look for quotes from Google if you've actually done it yourself. I've managed THOUSANDS of EMDs. They took a hit when Matt Cutts made a comment about EMDs taking a hit. Now here you are in lala land, saying there was never an EMD bonus to begin with, because Google never said there was one?

You've demonstrated you completely have no idea what you're talking about.

 

[Scritty]

I completely agree with you 100% on this.They also apply statistical models for their anti-spam algorithm on youtube comments, especially for determining a thumbs up or down being valid or invalid.

Google algorithms are heavily centralized around "Ensemble Trees".To be honest, I would suggest anyone that is serious into using the right scientific way to break their algorithm or atleast come close to it, to actually collect a bunch of data and use statistical data compressed models.



The paper your describing is pretty much using clusters of computers put together,sort of like Hadoop but using their own , MapReduce

To the rest believing that fairly tale shit, get real.

Google has mathmaticians using statistics and state
of the art tree learning algorithms.

By the way, that paper is pretty much one of the most famous papers from a google engineer. It is pretty old, but its extremely popular even among non-academic people.

This is more like it. ^^^^^^^^^^

"Failure to click" is a great idea in principle...if it worked as is "Ignoring the top result on purpose". But in practice those two ideas are only a factor in the ranking algorithm, and to get substantial traction on a site by gaming the system like this would take one hell of an effort and offer zero guaranteed results.

All that effort could just be spent making your own site a little better.

Scritty

EMD's DO STILL HAVE AN EFFECT! Sometimes only in minor niches sometimes only short term, but a quantifiable effect all the same.
Google reduced the effect of EMD's last year, but did not remove it completely.
For shit sites this reduction was enough to remove any chance they had in ranking high
However, a bigger negative slap last year at the same time as EMD leveling was the over commercialization hit. Many MFA sites with EMD's had adsense blocks very high on the page and while the webmasters attributed the fall in serps they had to EMD leveling, in many cases it was due to over commercialization slap.

Correlation does not equal causation

 

[grahamklemm]

TL;DR

But I will say this from glancing briefly.

Here you are again above, trying to look for a quote from someone at Google saying there was a bonus for having an exact match domain in the last few years.

Really, WTF? Haven't you even tried? Why do you even need to look for quotes from Google if you've actually done it yourself. I've managed THOUSANDS of EMDs. They took a hit when Matt Cutts made a comment about EMDs taking a hit. Now here you are in lala land, saying there was never an EMD bonus to begin with, because Google never said there was one?

You've demonstrated you completely have no idea what you're talking about.

I never said there was NEVER a bones for EMD...in fact there most certainly was a bonus for it in the early days of search engines and I exploited that. But that loop hole closed years ago which is why I stated no-one from google in the last few years has said there was a bonus and clearly from experience and observation there is not one. Im even surprised that anyone would think such a thing. It shows a complete lack of understanding of everything. Your understanding of search engines is very simplistic...sort of kindergarten stuff.