WH info from a BH mind

[ipopbb]

interesting. So are we saying that the page should have a blockquote and an address tag?

I think the jury is still out on what this all means, but blockquote and address tags are looking like the power tags to me too.


I am really interested to see what Spaceman comes up with. The notion of second tier relationship testing to find out if 3 C tests = 1 A Test seems like a very important next step. If keyword usage in certain kinds of content is a currency then we need to know... It might also be a max usage or a first usage or even an average usage... we just don't know but the notion being discussed is that we now might have an empirical tool to test with and reverse engineer the "content score" piece of the google equation.

 

[ipopbb]

The factors that I believe are becoming more important are:

Domain Age
Domain Name e.g. keywordphrase.com
Offsite factors & Link building

Domain age isn't quite right... It seem like it is but its not. It is actually the age from when google first indexed it that matters. I have access to about 700 old domain names going as far back as 1990 that have never been used. They get no bonus over a new domain name. However... if google has had the domain indexed for a long time... then it gets lots of rep.

A good way to validate an old domain is worth something is to see if wayback machine has entries for it. Old domain with no entries to match is worthless.

http://web.archive.org/web/*/http://www.blackhatworld.com


Its a subtle distinction but google trusts old websites. An old domain that only recently gets a website is a new website in google's mind.

 

[stevesdata]

Domain age isn't quite right... It seem like it is but its not. It is actually the age from when google first indexed it that matters. I have access to about 700 old domain names going as far back as 1990 that have never been used. They get no bonus over a new domain name. However... if google has had the domain indexed for a long time... then it gets lots of rep.

A good way to validate an old domain is worth something is to see if wayback machine has entries for it. Old domain with no entries to match is worthless.

http://web.archive.org/web/*/http://www.blackhatworld.com


Its a subtle distinction but google trusts old websites. An old domain that only recently gets a website is a new website in google's mind.

Yes this is what I thought. I think the longer g00gle has indexed it is very key. A client I am working with atm has a completely bs site but very old and its PR5 with no SEO at all (not even the most basic stuff I can think of, like really unprofessional lack of html tags).

I still believe an older domain name is far better than a shiny new one though.

There was quite a smart chap that posted he created sites and left untouched for long periods and then developed later and saw good results.

 

[ipopbb]

There was quite a smart chap that posted he created sites and left untouched for long periods and then developed later and saw good results.

I do this too, it works... the magic wait time seems to be about 10 months but can vary + or - 2 months based on where you fall in Google's propagation cycles.

There is probably a business to be had in selling properly aged domains... I can see people paying a premium for the right now effect.

 

[ipopbb]

Summary: A week of surprises in reverse engineering Google's content score. Title tag makes a huge rebound. Alt text and link text fall out of results?!?!

Theory: link text and img alt text apply credit to src and href not document. I should make an empty page and image on another unused host and see if they rank for the keyword instead of the originating pages.

Top Keyword Usages

1. Code 1: Keywords in title tags
2. Code 14: Keywords in blockquote tags
3. Code 55: Keywords in dl,dt,dd tags
4. Code 17: Keywords in address tags
5. Code 11: Keywords in p tags
6. Code 51: Keywords in td tags inside a tfoot block
7. Code 25: Keywords in dfn tags
8. Code 7: Keywords in underline tags
9. Code 62: Keywords in h6 tags
10. Code 21: Keywords in caption tags
11. Code 47: Keywords in sub tags
12. Code 29: Keywords in var tags
13. Code 46: Keywords in strike tags
14. Code 6: Keywords in b tags
15. Code 3: Keywords in h1 tags
16. Code 27: Keywords in samp tags
17. Code 50: Keywords in td tags inside a thead block
18. Code 8: Keywords in i tags
19. Code 34: Keywords in ins tags
20. Code 12: Keywords in span tags
21. Code 22: Keywords in center tags
22. Code 61: Keywords in h5 tags
23. Code 35: Keywords in label tags
24. Code 44: Keywords in s tags
25. Code 2: Keywords in visible body text
26. Code 5: Keywords in h3 tags
27. Code 24: Keywords in del tags
28. Code 26: Keywords in em tags
29. Code 49: Keywords in td tag
30. Code 43: Keywords in q tags
31. Code 15: keywords in abbr tags
32. Code 31: Keywords in font tags
33. Code 58: Keywords in caption tags
34. Code 28: Keywords in input tags
35. Code 9: Keywords in strong tags
36. Code 42: Keywords in pre tags
37. Code 4: Keywords in h2 tags
38. Code 36: Keywords in li tags inside ul tags
39. Code 18: Keywords in tt tags
40. Code index: A page that links to all tests
41. Code 40: Keywords in select, option tags
42. Code 41: Keywords in Object param tag name attribute
43. Code 37: Keywords in meta description
44. Code document: Keywords in file name
45. Code 52: Keywords in HTML comments
46. Code 39: Keywords in meta author
47. Code 38: Keywords in meta keywords
48. Code 32: Keywords in input tag value attribute
49. Code 57: Keywords in a tag href attribute

Fell out of Results

• Code 56: Keywords in a tag link text (formerly 7th)
• Code 53: Keywords in img alt text (formerly 8th)

Yet to appear in Results

• Code 10: Keywords in cite tags
• Code 13: Keywords in div tags
• Code 16: Keywords in acronym tags
• Code 19: Keywords in big tags
• Code 20: Keywords in button tags
• Code 23: Keywords in code tags
• Code 30: Keywords in legend tags
• Code 33: Keywords in textarea tags
• Code 45: Keywords in small tags
• Code 48: Keywords in sup tags
• Code 54: Keywords in in dl, dt tags
• Code 59: Keywords in bdo tags
• Code 60: Keywords in h4 tags

 

[pharmokan]

skewed results from an overly elaborate algorithm that noone has yet to crack.

 

[LucasCuello]

Very helpful, congrats for your work, you must be the first person doing this! Will read the next posts!

 

[w4rl0rdx]

Yes, very cool. Great Job, Keep us updated. I'm curious to see what happens.

 

[squelsh]

How's this doing in Yahoo and Bing?

 

[me100rabh]

How's this doing in Yahoo and Bing?

ya u should also try it in other search engine. this helps you to refine the result.

 

[squelsh]

Another point on this. This is all great info but it has only been tested on one domain.

The work involved to test something like this is minimal so if we could get a group of people to run exactly the same test (different test word obviously) on their own domain this would be much more scientific.

I would be more than happy to buy and set up a .info and run the same test if others would as well. Obviously the more people that do this the more conclusive results.

Anyone else want to join in?

 

[dynander]

I will like to help with this experiment too. As well as squelsh, I can buy an info and run the same test with other keyword

 

[Albert]

Sounds like an awesome experiment ipopbb. Great to see someone actually doing the testing. Now this can be like culprit of what factors change over time in Google's algorithm.

 

[Silversten]

This is a great experiement but in order for us to get any meaningful results it needs to be replicated, perhaps 100 times.

Perhaps someone could design a script for this?

 

[divinci]

Lovely thread + thanks.

Gonna have a think of some spin off experiments

 

[niksanyl]

This is a really cool test. I have a few questions. All of these pages were created on the same domain correct? SInce these do not link to each other, how does google find the pages? did you create a sitemap? I'd love to experiment with this!
Thanks!
Niksa