Big changes in Google content scoring

frankeh said:

Basically I enter a keyword, it checks the monthly search volume. Then from there I would check the other stats as applicable. Surely that would then only be 1 call as I only want 1 stat back..

Love this and you other threads. Really informative.

Click to expand...

Look ups on a single keyword is 1 Unit to my knowledge.

MisterGemini said:

Rather than making cursory remarks about some elements going up while others going down on the data you are releasing, why not pop it all into a spread and generate a nice chart graph to show just how radical the moves might be?

At least this way you can get a historical view, and it might show some particular elements to be consistent in their import.

It will also help people understand just how many changes are there to consider from quarter to quarter.

This is a great share nonetheless. Good work.

Click to expand...

My followup to this thread has a movers & shakers style trending... I also have charts of common SEO factors in another thread too! I will try to provide more visuals in the future though for sure!

Thanks!

iglow said:

i tried to use <samp> and <cite> as u mention and still my old pages with <h1> and <strong> rank better than it

Click to expand...

Thanks for sharing your results. With the MayDay update google is fubar'd... It doesn't shock me at all that tweaking content right now does little or nothing... Its a whole new Google in my opinion... I have to repeat a lot of old experiments in the near future and change a lot of things on my current sites just to get back to where I was in early april.

Your findings are every bit as valid as mine. I gave your post a "thanks", because I respect and appreciate that you tried and then shared which is ultimately what gives any experiment legs.

I would recommend you try again in a few months when google sorts their shit out... if you still wonder... I understand though often one time is all an idea gets because there is no way to try them all.

ipobb can u give a code sample of a html page u think would rank the best for lets say 4 words keyword?
i think the infos u gave are valuable, i was also personally using some less known tags but i cant exactly track the gold here.

if the keywords is lets say:
buy dog food now

how would u build 1 page mostly optimized for this?
i do not mean the keyword dentistry - thats everybodys job, just tag wise.

iglow said:

ipobb can u give a code sample of a html page u think would rank the best for lets say 4 words keyword?
i think the infos u gave are valuable, i was also personally using some less known tags but i cant exactly track the gold here.

Click to expand...

I'll do better than that! I use this example a lot. It's not my site and if people screw them they are screwing cancer victims... so its a pretty safe example.

Search for Pink Ribbon
look for Pink Ribbon Shop .com

Google's algorithm right now has 4 factors + website age (not domain age... there is a difference)...

1. Absurd Keyword Density
2. Ok Page Rank
3. Dmoz & Yahoo directory Links
4. Keyword in domain (or URL if your not gunning for #1)

I don't know how long it will last but those factors appear to place 8/10 page 1 serps in most of my test cases.

I don't know why google did this... It is sooo exploitable. There are three theories that I think have legs right now.

1. Google fucked up and the MayDay/Caffeine deployment was too big to roll back. This is the Google we have going forward and they will make tweaks and adjustments over time going forward to correct the SERPs eventually.

2. This is a shareholder value play to make online businesses spend more in adwords instead of getting their search traffic for free organically. (This is exactly what everyone who lost their longtail traffic is doing. It might be a good time to invest in google, because with all the extra competition google's is going to make bank on this. Of course, I'm reluctant because I fear their P/E ratio. Too steep for me.)

3. This is a honeypot strategy to get black hats to self select themselves out of the SERPs.

I ordered them in order of which one I think is most likely the case... BUT... bing has the exact same SERP anomalies... It is totally freaking me out because how do they tune to the same quirks and failures unless they are colluding, which would lend support to theory 3 if other people see it too. So far its just me and a couple others...on the Bing stuff observations.

Also:

AddictingGames.com for the search Games
Seo.com for the search SEO

I like some of the strategies like using keywords in comments, js variable and methods, css class name, title and id attributes, if keywords are in the domain then use full URLs to everything links images includes etc... Anything the get the raw number of page matches up without scaring your visitors. My experiment in this thread does indicate that HTML comments and some attributes can rank a page for a keyword... otherwise they would not be in my test serps at all, let alone in the middle of the list in some cases! That is some empirical GOLD no matter what people may think of my theories!

I know my theories and conclusions are "out there", but my data is real... and it is what it is... data.

@ipopbb - I've read through your threads. Amazing information and theories. I personally really appreciate you taking the time to share this.

Google's algorithm right now has 4 factors + website age (not domain age... there is a difference)...

1. Absurd Keyword Density
2. Ok Page Rank
3. Dmoz & Yahoo directory Links
4. Keyword in domain (or URL if your not gunning for #1)

Click to expand...

DMOZ.. really?

How does the website age differ from the domain age? Is it the in favor of being new or old?

What is an OK PR? 3 maybe?

I agree about the keyword domain stuff, that hasn't really changed much at all over the years.

MisterGemini said:

DMOZ.. really?

How does the website age differ from the domain age? Is it the in favor of being new or old?

What is an OK PR? 3 maybe?

I agree about the keyword domain stuff, that hasn't really changed much at all over the years.

Click to expand...

Yes... dmoz... google directory is back in the game now too so a dmoz link = 1 from dmoz, 1 from GOOGLE, and 1 from every other directory that reposts the database... google link: pinkribbonshop.com from my example... They are PR3 site beating PR6-7 on page 1... what makes the difference for them... the other three ranking factors I mentioned above.

Google knows the date of when a website was first indexed by the search engine AND google ranks, rewards, and punishes by full hostnames. Not domains, which is why webmaster tools lets you control how google uses your hostname. I have plenty of domains that have never been used for websites from the mid 90's... they have no extra benefit over a new domain because they are both new to google. I also have domains that were used as websites in the late 90's and google remembers them and they have the aged website value. Google may have use domain created date in the beginning, but at this point in the game google has a better age factor... its own memory. (which is why you should use wayback machine to see how old the website is instead of using the age of the domain... it makes a big difference in many cases)

For my keyword space the domain keywords weren't a big factor at all until mid April this year (a few exceptions but major retail brands dominated). But my keywords were really big ones that most people would never even shoot for. Its all different now. Brands are decimated in the changes.

If you target mid to low competition keywords then you are probably right... domain matches probably always worked pretty well at that scale.

though this topic is never dead but thanks for sharing this.

ipopbb said:

Thanks for sharing your results. With the MayDay update google is fubar'd... It doesn't shock me at all that tweaking content right now does little or nothing... Its a whole new Google in my opinion... I have to repeat a lot of old experiments in the near future and change a lot of things on my current sites just to get back to where I was in early april.

Your findings are every bit as valid as mine. I gave your post a "thanks", because I respect and appreciate that you tried and then shared which is ultimately what gives any experiment legs.

I would recommend you try again in a few months when google sorts their shit out... if you still wonder... I understand though often one time is all an idea gets because there is no way to try them all.

Click to expand...

I did the exact same thing for one of my sites and 12 out of the 16 keywords I look for got worse including my main keyword which went from #7 to #11. Aside from the usual dance, that was the only possible negative change. (I'm still link building in the background via articles and profiles)

Extropy said:

I did the exact same thing for one of my sites and 12 out of the 16 keywords I look for got worse including my main keyword which went from #7 to #11. Aside from the usual dance, that was the only possible negative change. (I'm still link building in the background via articles and profiles)

Click to expand...

Thanks for sharing... can I ask what arena you play in? Adsense, affiliate, retail, etc...?

ipopbb said:

1. Absurd Keyword Density
2. Ok Page Rank
3. Dmoz & Yahoo directory Links
4. Keyword in domain (or URL if your not gunning for #1)

Click to expand...

I can imagine that the conclusion shouldnt be that google is favoring these factors so much but instead that these webmasters did the best seo for their sites.
Especially point 4. If a webmaster goes so far to get a domain that has the keyword inside its clear he wants to get this key on top. And because of that he probably will do a lot for that.
On the contrary webmaster with big sites create a page with this keyword but probably wont spend the same amount of work to get this page on top.

I think that the top websites include these 4 points only points out that they did everything they could to rank it. More than others. And if the others would do the same work without these 4 points they probably can compete with the sites that include these 4 factors.

Only guessing...

If you are serious with your v2, you'll be a very rich jerk soon.
In my 12 years of optimisation experience, I had five times access to such (similair) tools, gorgeos and highly effective as you describe - but they all disappeard very fast and the only way too explain this is - they have been bought by our big brother as it is too easy to achieve top rankings with such information.

ipopbb said:

I doubt that makes a difference... I already have some data along those lines involving URLs with query parameters in them... google doesn't mind and they trend upwards as you approach page one. If the mime-type is text/html google treats it all the same.

File types might rank differently for keywords... thats in my version 2 of the test suite. People have asked for my source code and I plan to deliver it in a version 2 of the experiment probably later this week. I just need to convert my report tool to PHP from java... I removed some stuff that was redundant or not producing and added some new tests and added more detail to others... The new test group will be 180ish test cases and will come with instructions to step it up properly.

By the way, I noticed then, they worked even better for Yahoo (years ago).


Cheers

Click to expand...

This is one of the most interesting threads on BHW i read in a long time.

We really need a practical way to APPLY this knowledge. Your list is really good, but i have to admit even as someone into HTML and PHP and making sites for years...i didnt even know about <samp> or <acronym> or <kbd> and quite a bunch of those tags!

Now i/we need a practical guide how to implement this....eg. how to mod a wordpress theme to use those tags or what we can *very real* do to our content to optimize it that way!

bh

Edit: You are referring to test-suite. Is this available somewhere?

really nice information!

its hardwork to get result like this

thanks

blackhaze said:

This is one of the most interesting threads on BHW i read in a long time.

We really need a practical way to APPLY this knowledge. Your list is really good, but i have to admit even as someone into HTML and PHP and making sites for years...i didnt even know about <samp> or <acronym> or <kbd> and quite a bunch of those tags!

Now i/we need a practical guide how to implement this....eg. how to mod a wordpress theme to use those tags or what we can *very real* do to our content to optimize it that way!

bh

Edit: You are referring to test-suite. Is this available somewhere?

Click to expand...

I was about to suggest the same thing. It would be great if this thread could grow beyond the 'wow this is great' comments, to perhaps actually doing some analysis on the various tags, what they are for, how they work, and best practices to use them. I believe this will go a long way towards bringing more meaning to others about how to use this information.

I think the OP is the best to start on this, but if others have experience with the other various tags, would certainly be good as well.

MisterGemini said:

I think the OP is the best to start on this, but if others have experience with the other various tags, would certainly be good as well.

Click to expand...

If only I had the time... If I get a chance in the near future I want to re-compile my Top N Trends charts for SEO factors... I think that data would be very useful to people in this post Google Caffeine/MayDay World.

I mention a few times in this thread how I use this data. Others have not had the same luck with it, but I think timing might have played a role with that.

FYI... I posted June data as a new thread

June 2010 Tag Rankings

Awesome.

Your link doesn't work.. so here it is for those who get a blank page like I did:

http://www.blackhatworld.com/blackhat-seo/white-hat-seo/205465-june-2010-tag-rankings.html

Did you find any "faster" way to get into DMOZ listing ?

deki33 said:

Did you find any "faster" way to get into DMOZ listing ?

Click to expand...

Nope... and if I ever find one I'm not sharing that!