Google's Spam Recognition Guide for Raters

Very interesting reading....

Google's Spam Recognition Guide for Raters.

Google's CommQuest tames biased raters.

Google's guidelines on random-query evaluation.

All from Search Bistro.

- Michael

 

Interesting. They mention Joeant (as a good one). Toxic Lemon gets hammered though and if I remember correctly, that's a DP member's site! (Doing a search proves me wrong, I must have seen it somewhere else...)

This all seems a bit unprofessionaly written though...

 

WOw! This is great info!

 

I guess a lot of brightsparks will go "Ow *scratching head*, so that's why my site got de-indexed...".

 

Thanks, Michael!

 

Great info Micheal.

Just imagine the how your stomach would sink through the floor if you opened that doc to find your site mentioned as one of the baddies. I clicked through on one that was quoted for hidden text (the first one) and didn´t find any. Have the cleaned house already?

 

damn google is going to be ruining affiliate money.

 

I was amazed by how unsophisticated those documents were.

That is all Google is looking for as spam?

Wow.

Black hats of the world rejoice!

 

Very intereseting. Good job, Michael

 

Great info. What it boils down to is that Google is willing to pay actual people to determine if a site adds value for the user. Which is what google has been saying all along.

 

Some were dated 2003. Could be close to 3 years old.

Has this Bistro dude been threatened with lawsuits yet?

 

Well that's my sites fu#^$d. They are all affiliate sites....it actually makes me feel like a criminal, I'm waiting for the knock on the door with the Google police coming to take me away.

I actually find most of the that article offensive, particularly that pertaining to affiliate sites. Thousands of people buy through affiliate sites every day.....it's like Google is saying everyone has to buy from the one department store, all the others are illegal.

It's big brother gone mad

 

100% Frame - Hmm.. I wrote an affiliate link protector which uses a 100% frame to render an affiliate site. But I developed the tool specifically for assisting people to prevent affiliate ID hijacking. It had nothing to do with manipulating Google. In fact, the links work through Javascript, so the affiliate link frame pages shouldn't even appear in Google's index. This document/guide doesn't mention that there are legitimate uses for a 100% frame.

I suppose this speaks to sites that republish press releases, articles, RSS news feeds, scraped search results, etc. IMO, this is a good thing.

Re: Thin affiliates - Looks like Noppid's FreeBS pages would be labeled as thin affiliate pages.

 

I agree with Google on most stuff.

Quite frankly Cyclops, if your site does NOT offer any added value and you can order the same stuff from the real store then WHY should they keep a copy of your site in the index?

Suppose it's a popular affiliate program amongst SEO's (like Amazon Books as Bernard points out) so the top 50 gets saturated with well optimized results for this ONE RESOURCE. The searcher will get pretty much 50 similar results of which at least 49 are obstructing other potentially useful results.

Be First or Be Different.

If you don't ADD value, don't cry if they drop you.

Solution could be you personally write an honest review of all products. Or you also promote the competitor affiliate's products and offer a price comparison aling the way.

The less copies in the index, the better IMO. It can be a nightmare finding stuff nowadays.

 

thanks for the great info!

 

I know of one national brand named mass merchant that isn't going to be happy about this info. I have seen no less than 20 sites (affiliates) with exactly the same products, descriptions... you name it. The do have other avenues to pursue however and I doubt they will be (are) adversly effected to any great extent. Especially since it seems that this has been inplace for quite some time. At least Google is stepping up to the plate. It gives my sites a better chance at gaining page one. I'm sure I would have a different opinion if I was in the affiliate biz.

I agree that if an affiliate site offers added value it's a different story but most that I've seen don't.

 

Very very interesting.

They say that a site offering people to join it's affilliate program can't therefore be an affilliate itself. Isn't there aff programs that are multi tiered?

I wouldn't be unhappy if you rely heavily on affilliate income, that just detailed how to avoid a negative rating. Sounds like a hell of a lot of work if you have a lot of products, but adding value will get you through their hoops.

 

Interesting reading. Thanks for the links.

 

In the case of affiliates it's all about moving them over to adwords, it's as simple as that.

I don't know where the line is drawn with added value, my main affiliate sites offer a range of products with descriptions and gives the customer the choice between which one best suits them.

Even Noppids amazon script gives people the option to choose which book best suits them for the topic they are searching for, the range of books is so vast that it does give added value , so yes it is unique.

I admit two of my sites are crap, I just placed affiliate banners on them until I get around to working on them but even so, they are still one stop shops with various products to choose from. I think if you offer people a choice between products that it is added value.....one search instead of ten.

If you follow Googles logic there would be no place for corner stores, just one gigantic chain of mega-marts.

 

Dunno if this has been mentioned elsewhere, but I was wondering whether G is using its humans reviewers to create a giant Bayesian filter?

I am not a techno kinda guy by a long way, but as I understand it, a Bayesian filter basically "learns" statistical patterns from user inputs. For example, I use a free sp@m filter called Spamihilator that has this feature. I classify emails as "good" or "evil", and gradually it becomes "smarter" and can figure out whether a completely new email is sp@m or not, even without having seen it before.

Although the reviewers would not be changing the algorithm directly, their choices and ratings could be used to create a similar Bayesian filter layered on top of the base algorithm. If that is the case, it could result in a great improvement in SERPs quality, if done in a fair and well-controlled way.