SEO Lab: Tag Ranking - HTML source for test cases

Extropy said:

Thank you for sharing this with us. I plan on making some tweaks to my main site's pages based upon your results.

If you don't mind zipping up your test files, I'll be happy to change the nonsense word and post them up on one or two of my sites.


Justin

Click to expand...

I've run very short on time unfortunately and I want to keep my promise of providing source... I haven't had a chance to rewrite the reporting tool in PHP. But that's the nice to have part... not the essential part.

I left off a little more work at the end of the test cases... I wanted to add different document types using the KEYWORD to see how they fair...

These are my version two test cases... They will provide better understanding than the current test cases because they are more detailed where the need to be and less detailed where it doesn't matter...

Good stuff, thanks for sharing this with the community! Time to get testin'

I have no idea what these are used for. How can I use them ?

clifford said:

I have no idea what these are used for. How can I use them ?

Click to expand...

These are for folks who have been following my threads and want to try my experiment for themselves.

Big changes in Google content scoring
Follow-up to Google Content Scoring Changes
More Empirical Data From MY SEO Lab
SEO Lab: PHP Source for proxying Google searches
SEO Lab: Tag Ranking - HTML source for test cases

So I throw in my nonsense keyword, and load it onto an empty domain?

sevenfigurez said:

So I throw in my nonsense keyword, and load it onto an empty domain?

Click to expand...

Yeah... and perform no SEO on the pages or domain... then submit the index page to google and submit the url test cases to google independently because they cant be in the index page. Then wait a couple weeks and start querying the imaginary word and collect the order of the test codes from the SERPs.

Track changes and show in a movers and shakers report a.k.a up 5 down 3 etc...

Awesome, thanks man

- Actually this part I am not understanding: "then submit the index page to google and submit the url test cases to google independently because they cant be in the index page."

Are the url test cases you are referring to 000, 001, 002, 003, etc.?

sevenfigurez said:

Awesome, thanks man

- Actually this part I am not understanding: "then submit the index page to google and submit the url test cases to google independently because they cant be in the index page."

Are the url test cases you are referring to 000, 001, 002, 003, etc.?

Click to expand...

My apologies for being scatter-brained... I didn't realize I did the separation for you... The URL test cases are in index2.html... submit index.html and index2.html separately.

Been looking forward to this for awhile, going to do some rich snippets testing, too. Thanks!

EDIT: I find it amazing that a lot of bogus WSOs get more attention than your posts. This is the most useful information I've seen on BHW.

This must be on new empty domain or I can just add it in one folder on domain where is existing site ?
Also links on index.html to all tag pages are in some order - do you think google gives more value to pages (tags in this case) that are more on top of index page.

For keyword - do you use one word keyword like "ds9d9sadk7d54" or some long tail keyword that doesn't exist anywhere else like "kisa slaok fsai dasko"

I just uploaded this to 5 domains. They are all different in age, PR...
I will use this in folder and link to index and index2 for main page of site to get crawlers faster.
Can you share some tool to faster compare results after few weeks ?

Now this is some amazing work.

deki33 said:

This must be on new empty domain or I can just add it in one folder on domain where is existing site ?
Also links on index.html to all tag pages are in some order - do you think google gives more value to pages (tags in this case) that are more on top of index page.

For keyword - do you use one word keyword like "ds9d9sadk7d54" or some long tail keyword that doesn't exist anywhere else like "kisa slaok fsai dasko"

Click to expand...

You got the idea... I wouldn't say any one config is any better than another but if you think it through you'll find that how you stage your experiments can limit the kinds of tests you can conduct.

I have only used single imaginary words with not punctuation, whitespace, or digits. That doesn't mean you have to. In fact it might be more interesting if you did try something different.

I'll let you decide.

deki33 said:

I just uploaded this to 5 domains. They are all different in age, PR...
I will use this in folder and link to index and index2 for main page of site to get crawlers faster.
Can you share some tool to faster compare results after few weeks ?

Click to expand...

I will upload my java source... perhaps someone can rewrite it as a php file to package with the test sources.

Give me a couple days. I have to pluck it out of some more sensitive code I can't share.

From 5 domain I can see one on google with 4 pages indexed:
133
048
keyword in folder name
013

takes time... Try a site search to see if you can pull more at the moment.

also... the test pages look very similar to google because they are so small. You will get this message at the bottom

In order to show you the most relevant results, we have omitted some entries very similar to the 6 already displayed.
If you like, you can repeat the search with the omitted results included.

Click to expand...

To avoid this add &filter=0 to your search URL... I also like to add &num=100 too.

Thank so much for sharing this... quick question, does the domain need to be different from the keyword? ie if my domain is fdlskeg.com then should my be fdlskeg or something different like walskrhe or does it not matter.

thanks, Joolez

joolez said:

Thank so much for sharing this... quick question, does the domain need to be different from the keyword? ie if my domain is fdlskeg.com then should my be fdlskeg or something different like walskrhe or does it not matter.

thanks, Joolez

Click to expand...

yes ... with 1 exception... the test for keyword in the domain name.

also need a 2nd domain if you to test for keyword in the CNAME i.e. keyword.2ndDomain.com

If you place them all in the secret domain... then you lose the ability to do the domain and CNAME tests, but the relative rankings are all still valid because they all have the added domain match which makes them sort by their differences still.

FYI... for the people following my threads... I shared with Joolez in a PM that the community is very divided still on whether or not these measurements mean anything at all. Obviously people know where my optimism stands, but I recommend anyone trying this experiment keep an open mind the other way too. Being proven wrong is fine with me. It frees up my time and money to work on other equally exciting experiments. (I just need proof because my 11 month stretch of Tweak & Gains was rock solid to me.) Its a different Google now though. Tweak & Gain stop working for me...largely because Caffeine/MayDay changed the Algo so drastically. I'm still watching this data. I'm hoping to retune for Caffeine and then recoup the "fine tuning" tactics this originally provided me. It could just be dead now. I just won't know until I'm back where I was in the SERPs... I takes me 4 days to pull the site rankings for all my keywords & sites... I am still doing really well in Short Tail... What I've lost is the 500K long tail keywords I used to rank for. My first broad scale attempt to correct this will deploy next week. I expect to see large traffic and revenue gains from it... If I do I will share the vital aspects (probably not all the details) but enough for others to correct as well.

Cheers

Thanks Ipopbb for your help... I'm interested in the experiment too! And always up for a bit of tweaking. My seo lab is now live, so I'll check in once I start getting some results

Thanks for shereing with the rest of us man...