Search Engine Watch
SEO News

Go Back   Search Engine Watch Forums > Search Engines & Directories > Google > Google Web Search
FAQ Members List Calendar Forum Search Today's Posts Mark Forums Read

Reply
 
Thread Tools
Old 12-22-2005   #21
DaveN
 
DaveN's Avatar
 
Join Date: Jun 2004
Location: North Yorkshire
Posts: 434
DaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to all
Andy ... was that a "I can and can not confirm the exisitance of an effect that some people may or may not call the sandbox"

DaveN
DaveN is offline   Reply With Quote
Old 12-22-2005   #22
jkemp
Member
 
Join Date: Dec 2005
Location: Tennessee
Posts: 5
jkemp is on a distinguished road
All I can say is what I have seen from personal experience. The company I work for launched a web site for a very high profile hotel that was being constructed in downtown Knoxville, TN, http://www.cumberlandhousehotel.com. The site was launched in late summer. Because it was so high profile, it gained a lot of links locally very quickly. A few months after launch, the Google Toolbar showed it had a PR of 4, but it did not show up in the results even when searching for the hotel name plus the location. If the sandbox doesn't exist, how do you explain that?

Google has to have some kind of filter on domains that are a certain age. I believe in this case that the site got too many links too fast in addition to being new and Google deemed them to be unnatural. How do you explain that to a client? I would consider that good marketing, advertising, pr, whatever. Now granted, the site has almost no content, which I explained to them was not good. I think this could explain why it still does not appear in many searches even though it shows up now when you search for the name although it is still like 3 or 4 down. I suggested to them to use Google Adwords and after they saw they got no traffic, they were convinced it was worth it. Now most of their search traffic comes from Google ads. Isn't that what Google wants anyway?

Last edited by Marcia : 05-28-2007 at 08:52 PM.
jkemp is offline   Reply With Quote
Old 12-22-2005   #23
DaveN
 
DaveN's Avatar
 
Join Date: Jun 2004
Location: North Yorkshire
Posts: 434
DaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to all
Jkemp thats something different .. sorry

DaveN
DaveN is offline   Reply With Quote
Old 12-22-2005   #24
AndyBeal
Fortune Interactive
 
Join Date: Jun 2004
Posts: 29
AndyBeal is on a distinguished road
Quote:
Originally Posted by DaveN
Andy ... was that a "I can and can not confirm the exisitance of an effect that some people may or may not call the sandbox"
Hah..so sitting on the fence doesn't work around these parts?

Is there a "sandbox" for new sites? Yes!
Does every new site have to endure it? No.
Do I know what criteria is used to judge which sites get put in the sandbox? No
AndyBeal is offline   Reply With Quote
Old 12-22-2005   #25
DaveN
 
DaveN's Avatar
 
Join Date: Jun 2004
Location: North Yorkshire
Posts: 434
DaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to all
Do I know what criteria is used to judge which sites get put in the sandbox? No

I will stand beside you on that one too !!

DaveN
DaveN is offline   Reply With Quote
Old 12-22-2005   #26
DaveN
 
DaveN's Avatar
 
Join Date: Jun 2004
Location: North Yorkshire
Posts: 434
DaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to all
I posted this a few week ago .. still going strong the domain was brand new !!

brand new site launch 6 weeks ago . last weeks logs

1 No Referrer 5,296
2 http://www.google.co.uk 3,103
3 http://www.google.com 1,580
4 #*$!xxxxxxxxxxxxxxxxxxxx 901
5 xxxxxxxxxxxxxxxxxxxxxxx 714
6 http://search.msn.co.uk 336
7 http://www.google.com.au 120
8 http://www.google.ca 83
9 http://aolsearch.aol.co.uk 79
10 xxxxxxxxxxxxxxxxxxxxx78
11 http://www.google.nl 76
12 http://www.google.fr 66
13 http://www.google.se 57
14 http://www.google.de 53
15 http://www.google.it 44
16 http://search.msn.com 40

brand new domain ... xxxxxxxxx are sites that you will need to find imo

DaveN
DaveN is offline   Reply With Quote
Old 12-22-2005   #27
bewarne
Newbie
 
Join Date: Jun 2004
Posts: 3
bewarne is on a distinguished road
For what it is worth, here is my experience with the sandbox/age/credibility thing at Google:

1) Launched site about the TV show "House" (http://www.housemd-guide.com) on May 13.
2) Submitted to Google, Yahoo and MSN at that time and got links from several sites about the show (Internet Movie Database, TV.com, all the other sites about the show or the stars). Got a link from Open Directory within a month or so. Also linked from my highly rated "West Wing" site. And the site started being mentioned in the forums.
3) All three search engines had the site listed in their databases within a few weeks.
4) And for Yahoo and MSN, the site was coming up for competitive words soon thereafter: Words like "House MD" or "House Fox TV Show". Was #1 or #2 iin MSN for several major keywords within a couple of months.
5) Google would have the site come up for anything that my site was just about the only site covering/using those words in any kind of decent density way and was just about the only site dealing with those words, subjects.
6) At 5.5 months (just short of the usual six months I had been hearing about), Google started listing my site on the first page for everything connected with the show. No major change to the site happened at that time.

What this indicates --- I do not think it proves anything--- is that you can come up in Google immediately for non-competitive words (your name, or information that you have that few other sites have) but for anything competitive, you at first end up last.

The speed with which the site got out of the sandbox effect, may be due to its informative instead of commercial nature.

Last edited by bewarne : 12-22-2005 at 04:15 PM.
bewarne is offline   Reply With Quote
Old 12-22-2005   #28
Jill Whalen
SEO Consulting
 
Join Date: Jul 2004
Posts: 650
Jill Whalen is just really niceJill Whalen is just really niceJill Whalen is just really niceJill Whalen is just really niceJill Whalen is just really nice
Quote:
What this indicates --- I do not think it proves anything--- is that you can come up in Google immediately for non-competitive words (your name, or information that you have that few other sites have) but for anything competitive, you at first end up last.
Yep, those are the exact same symptoms reported by thousands of others who've created new sites. Very standard.
Jill Whalen is offline   Reply With Quote
Old 12-22-2005   #29
Marshall Clark
Member
 
Join Date: Jun 2004
Posts: 23
Marshall Clark will become famous soon enough
Have to put my $.015 in on this.

From personal experience I see evidence for a number of hoops that new sites are required to jump through in order to develop competitive rankings. Older sites may already innately satisfy these requirements and not be aware of their existence.

Call it a “sandbox”, “a series of complex filters”, or “TrustRank” – when you get down to it’s just semantics (sorry)

The lack of a common vector in the SEO community regarding the cause of the “sandbox” might suggest a number of factors being analyzed in combination with only a portion of them having to be satisfied at any one time.

Wild speculation conclusively demonstrates that sandbox criteria may include:
* Age of the domain
* Age of inbound links
* Reputation of linking sites (and their links, etc.)
* Age of inbound links of linking sites (and their links,...ad infinitum)
* Presence of an inbound link/301 from an established, parent company (likely one with few other outbound links)

Best way to avoid the sandbox in client websites:
* Only take on Fortune 500 clients
Marshall Clark is offline   Reply With Quote
Old 12-22-2005   #30
JasonD
Member
 
Join Date: Jan 2005
Posts: 51
JasonD will become famous soon enoughJasonD will become famous soon enough
A damn fine research document!!!

Disclaimer:There may or may not be a reason or why the above URL has been placed in a thread that may or may not be discussing an effect that may or may not have a name that is related to a container that may or may not contain silicon based granules
JasonD is offline   Reply With Quote
Old 12-22-2005   #31
graywolf
Member
 
Join Date: Jul 2004
Posts: 13
graywolf will become famous soon enough
I have two sites each with half a dozen pages about the same subject. Both have lots of low quality inbound links from SEO owned sites with optimal exact anchor text, both are nowhere in the top 1000.

I picked up an old site from 1999 put up one page with one paragraph about the exact same topic. No external links with the desired anchor text. The only link is internal and comes straight off the home page. Within one week of putting the page up it went to number 15. The site doesn't have lots of high quality inbounds but does have links from a wide variety of IP's none of which are from SEO sites.
graywolf is offline   Reply With Quote
Old 12-22-2005   #32
randfish
Member
 
Join Date: Sep 2004
Location: Seattle, WA
Posts: 436
randfish is a name known to allrandfish is a name known to allrandfish is a name known to allrandfish is a name known to allrandfish is a name known to allrandfish is a name known to all
Lightbulb

Mike, you know I'm a big fan, but I do have a very good sandbox example - I believe this is exactly what we're all talking about when we say sandbox:

www.seobythesea.com
Registrar: TUCOWS INC.
Whois Server: whois.opensrs.net
Creation Date: 19-jun-2005
Expiration Date: 19-jun-2006
Registered by Bill Slawski (bragadocchio)

Google - "seo by the sea" - http://www.google.com/search?q=%22seo+by+the+sea%22

- 15,800 results
- ranks #45
- unique phrase that didn't exist prior to Bill's creation of the site
- every one of the 44 pages ranking in front of it (and hundreds behind it) link to the site
- Yahoo! and MSN it's #1
- 1,000+ links according to Yahoo! linkdomain from lots of respected, on-topic industry sites

that's what I call being boxed. Bill's never bought unnatural links, his site deserves recognition and it's been getting it, naturally! I've seen this trend many times - seomoz.org itself was boxed for 9 months and couldn't be found for a search for "seomoz". Through the sandbox detection tool I've seen sites "boxed" even worse (although I can't confirm their link structure).

This isn't the exception to your rule, it's very common. And yes, big sites have been boxed, too; Electoral-Vote.com couldn't be found in the top 100 for 6+ months prior to the US election, despite being one of the most popular sites during that time (mentioned on CNN, CBS, Fox, every online news source, etc. - 100K+ links).

I'm just saying...

BTW
Quote:
Only take on Fortune 500 clients
This might be why you're not seeing it

Last edited by randfish : 12-22-2005 at 05:55 PM.
randfish is offline   Reply With Quote
Old 12-22-2005   #33
Marcia
 
Marcia's Avatar
 
Join Date: Jun 2004
Location: Los Angeles, CA
Posts: 5,476
Marcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond reputeMarcia has a reputation beyond repute
Quote:
ranks #45
#44 & 45, but not the homepage, which is nowhere to be found.
Marcia is offline   Reply With Quote
Old 12-22-2005   #34
jasonsot
 
Posts: n/a
Quote:
Originally Posted by randfish
Creation Date: 19-jun-2005
Expiration Date: 19-jun-2006

I agree 100% but wouldnt this leave a hole in this example

Last edited by jasonsot : 12-22-2005 at 08:10 PM. Reason: clarity
  Reply With Quote
Old 12-22-2005   #35
stuntdubl
Traffic not SEO.
 
Join Date: Jun 2004
Location: Upstate NY
Posts: 45
stuntdubl is a splendid one to beholdstuntdubl is a splendid one to beholdstuntdubl is a splendid one to beholdstuntdubl is a splendid one to beholdstuntdubl is a splendid one to beholdstuntdubl is a splendid one to beholdstuntdubl is a splendid one to behold
Post Give it a name. It's time, trust, and traffic.

Call it a sandbox. Call it a trustbox. Call it bad marketing. Call it poor seo.
"The Google Sandbox" has taken on a connotation of it's own depending on the vernacular by which which you frame your overall SEO philosophy and what you've personally ascribed to it's meaning through the interpretation of others (some who claim it doesn't even exist!) or your own personal experience (which for most is with a extremely limited data set of information).

The question of a sandbox brings up interesting discussions on other valid and important points of order for search marketers. I call it intelligent application of information from toolbar data, personalized search, and other applications which they've acquired over the last few years. It's the evolution beyond singular variables and data and storage issues. The irony is G is now sometimes not nearly as smart as we give the engine credit for, and sometimes it's way more intelligent than we possibly believe it can be. The "natural search algorithm" is an evolving organism with increasingly complex characteristics.

There are definite phenomenon related to time, trust, and traffic. Each of these are quality indicators that help sites achieve their ultimate goals. Both engines and webmasters.

Time
I think there is a substantial disconnect on age filtering right now, and it is an error that will be remedied on the engine's part in the next year. Not so much logical flaws, as lack of good philosophy. relevance = good enough right now rather than as good as it could be if applied more liberally and in different areas. The engines have found new and better ways to store and retrieve LARGE portions of data in the last three years. They are still figuring out how to disseminate it.

The new filters and ranking algorithms are much more sophisticated than hundreds of lightly dependant variables. There are literally multitudes of inter-dependant criteria that help to determine if a site or page is the "best". It is not the marketing success breakthrough for AI yet, but it is good enough.

Trust
The will get better at using personalized data, and acquire much more user data. If Web 2.0 is about user generated content, than the engines are WAAAY ahead of the curve. They are the kings of scraper content. They just haven't applied heavily because they are still contemplating long term issues of privacy concerns.

Traffic.
Fortune 500 companies don't get sandboxed, litterboxed, trust boxed, filtered, delayed, and seldom penalized. As long as the media likes it it's okay now I would think. If you're backing a site up with a TV campaign, you're not going to have a problem. The barrier to entry has been raised, and it's in the form of traffic validation of natural algo search results. The linking rich get filthy stinkin' richer, and those getting traffic amaze more virally. Think viral optimization - one of the nice recommended ways of beating the box.


There's only a few ways to learn about algos or conduct yourself as a SEO:
(I'd love to hear others chime in on how they learn SEO/SEM)
-listen and read from smart people who you trust -
make up your mind what's right and apply to sites
-listen and read from smart people who you trust
believe everything and don't test
-listen and read from smart people who you trust
test everything anyways.
-read whitepapers all day and try to understand their meaning and how they may be applied and spend time testing
-conduct your own scientific research against the efforts of dozens of well bread P.h.d's
-believe everything your read, and post wild theories to forums hoping that your competitors or their associate read and apply your disinformation
-blame everything on things like the sandbox and read blogs and check statistics all day.

Success = build quality sites and trust over time. It works for business and it works for goog.

MODERATOR NOTE: I've split the good stuff about what 2006 might bring into a new thread, see as how I'm in a thread splitting type of mood this week! You'll find it here, SEO/SEM Tactics For 2006, and I invite you all to explore non-sandbox stuff in that one.

Last edited by dannysullivan : 12-23-2005 at 07:53 AM.
stuntdubl is offline   Reply With Quote
Old 12-23-2005   #36
orion
 
orion's Avatar
 
Join Date: Jun 2004
Posts: 1,044
orion is a splendid one to beholdorion is a splendid one to beholdorion is a splendid one to beholdorion is a splendid one to beholdorion is a splendid one to beholdorion is a splendid one to behold
Exclamation

Well, hope this shed some light to the subject.


Last year I was discussing via email with Prof. Ricardo Baeza-Yates about web dynamics and temporal link analysis. He then sent me an email (Friday, November 12, 2004) in which he kindly attached a great work he published in 2002 with Felipe Saint-Jean and Carlos Castillo. Their work expands on the famous AltaVista's Bowtie Theory (a paper I read back in 2000)


Baeza's paper was: Web Structure, Age and Page Quality


The abstract states (emphasis added):

"This paper is aimed at the study of quantitative measures of the relation between Web structure, age, and quality of Web pages. Quality is studied from different link-based metrics and their relationship with the structure of the Web and the last modification time of a page. We show that, as expected, Pagerank is biased against new pages."


The authors tried to address a critical question. In their own words:

"Are link-based ranking schemes providing a fair score to newer pages? We found that the answer is “no” for Pagerank..."


When was this research conducted? Well, the data was collected from the last half of 2000 and first half of 2001. Note that this was way before SEOs started to promoting theories in discussion forums and SES conferences about a "sandbox".


Note the researchers mention that this was "as expected". So the phenomenon was related to the dynamics of the web and the way PageRank weight was passed through a dynamical system, not to any funny name "last updates from Google", a "new patent" or to some mythical "on/off page factors". ("If you do this or that to your site you can get sandboxed").


Although the study was limited to the chilean web and they defined age as "last modified", the age effect was well studied, addressed, and documented. The group later expanded their studies in two different occassion.

1. Web Dynamics, Structure, and Page Quality Baeza, et al.

2. Analysis of Link Based Ranking for the Web Baeza, et al.


In article 2 they wrote:

"What about the correlation between link ranking and age? Figure 8 shows the PageRank of all pages with respect to age. The bottom dots are normal pages, being the lower region, low ranked pages in low ranked sites, which is the most common case from the point of view of a link based ranking. The fact that most of the new or recently modified pages have low rank (the solid red region) shows that PageRank is biased to old pages. This is bad considering the constant change and fast growth of the Web."


In their concluding remarks they state:

"Considering that most visits are the product of a search, this dependency can have a large impact in electronic commerce as they benefit older sites."


Back then, these papers documented that for new pages there was an initial increment in page rank in the first 3 months followed by a drastic reduction several month laters in which the "true" PageRank is reached. So there was a period of many month (6 - 9) in which one would expect some sort of readjustment/oscillations in the scores. They also found that growth of the Web follows periods of bursts, all amounting to oscillations. These two papers are full of graphs which allows you to visualize the "age effect".


Is interesting to point out that after Baeza's work, Junghoo Cho and Robert E. Adams, from UCLA in

Page Quality: In Search of an Unbiased Web Ranking proposed a time-based model for page popularity and quality to solve the PageRank bias problem, but even in their model they found oscillations across several months.

Probably Google was aware of all those research findings and similar papers from those years and eventually decided to avoid all the hazzle by simply assigning a waiting period to new sites. To recognize these as new, probably they use a specific criterion (whois?, http headers?, last modified?, I don't know.)

Why the phenomenon was not reported before? Probably because back in 1998-99 no one studied the phenomenon or paid attention to it. At least I did not find any paper on the subject from 1998-99.

The fact that as of today unused aged domains can be utilized to try to game the system suggests that they may need to redefine better the "born on" stamp or whatever flag they might have before passing weights.


Orion

Last edited by orion : 12-23-2005 at 01:58 AM.
orion is offline   Reply With Quote
Old 12-23-2005   #37
gomer
Member
 
Join Date: Jul 2004
Posts: 13
gomer is on a distinguished road
Quote:
The debate is really about whether every new site will have problems ranking. The answer to that I think is firmly no.
I fully agree with this statement. But the ratio of sites experiencing this to ones that don't is what needs to be looked at.

I believe that well over 99% of sites launched will experience the sandbox. Yes, some sites getting links from whitehouse.gov and cnn can avoid this but that is like someone's analogy of winning the lottery or becoming president.

Would love to see a naysayer walk us through a real world example of getting a site out and ranking quickly.
gomer is offline   Reply With Quote
Old 12-23-2005   #38
DaveN
 
DaveN's Avatar
 
Join Date: Jun 2004
Location: North Yorkshire
Posts: 434
DaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to allDaveN is a name known to all
Randfish that example is different thats not the sandbox, it something else
DaveN
DaveN is offline   Reply With Quote
Old 12-23-2005   #39
rustybrick
 
rustybrick's Avatar
 
Join Date: Jun 2004
Location: New York, USA
Posts: 2,810
rustybrick has much to be proud ofrustybrick has much to be proud ofrustybrick has much to be proud ofrustybrick has much to be proud ofrustybrick has much to be proud ofrustybrick has much to be proud ofrustybrick has much to be proud ofrustybrick has much to be proud of
Quote:
Originally Posted by DaveN
Randfish that example is different thats not the sandbox, it something else
DaveN
DaveN, why not share with Randfish, what it is?
rustybrick is offline   Reply With Quote
Old 12-23-2005   #40
randfish
Member
 
Join Date: Sep 2004
Location: Seattle, WA
Posts: 436
randfish is a name known to allrandfish is a name known to allrandfish is a name known to allrandfish is a name known to allrandfish is a name known to allrandfish is a name known to all
Whatever it is, that's what I call sandbox. That's what the "sandbox detection tool" was made to look for and it's the phenomenon I've seen repeatedly.

Dave, Why must we have separate names for it?
and What's an example of what you call "sandboxing"?
randfish is offline   Reply With Quote
Reply


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off