Compilation of Anti-Sandbox Tactics

[siteseo]

While others, elsewhere, are discussing "is there a sandbox?" and "why won't Google admit it?" I think it much more practical to compile a list of theoretical remedies to the sandbox. Then maybe a few volunteers can test them and share their findings with other volunteers. So here's a first-pass at every theory I've ever heard on how to circumvent or shorten one's stay in Google's sandbox:
• Add a subdomain to a non-sandboxed site. First, make the site live on the subdomain. After it gets indexed, 301 it to the new URL. (I've also heard that this is a sure way to GET sandboxed, however)
• Send a press release via a service that feeds to Google News
• Get a link from a good, established web site (14+ months old)
• Create a private link network
• Develop a site in a language other than English
• Have NO outbound links from your site (not even in comment tags or images)
• Site is on unique IP address
• Get as many higher PageRank pages to link to your target pages as possible (older sites, created pre-sandbox)
• No inbound links from high PR sites at inception, then slowly add them
• Outbound links to "authority" sites
• Don’t have similarities to a “directory” (lots of OBL’s, no real “content”)
• Don’t do reciprocal links
• Differentiate IBL anchor text
• IBL’s should be deep-links, not just home-page links
• IBL’s from diverse IP ranges
• Yahoo Directory listing
• get links on pages according to PageRank value, rather than sheer numbers - or even topic - first.

Notice that some of these theories conflict with one another, and some can be combined into one process as well.

Please share your theories.

[kservik]

Originally Posted by siteseo

• Get a link from a good, established web site (14+ months old)
• Develop a site in a language other than English
• IBL’s should be deep-links, not just home-page links

Did all this for a Norwegian site (Kjøkkenguiden), but for the most competetive keyword (kjøkken) that I target I am heavily sandboxed (450-500, rank extremly well on other searchengines and other keywords).

Kim

[randfish]

Excellent compilation of ideas siteseo - thank you.

I can rule out several that have not worked thanks to my own experiences and testing. Since the site is already public in so many forums, you're welcome to see for yourself (it's avatarfinancial.com and the top terms are 'hard money' and 'bridge loans').

Your suggestions I've tried:
> Send a press release via a service that feeds to Google News
Have this
> Get a link from a good, established web site (14+ months old)
Have 100+ of these
> Site is on unique IP address
Have it on seperate C-block all alone
> Yahoo Directory listing
Have this
> Differentiate IBL anchor text
Have this
> IBL’s from diverse IP ranges
Only a couple shared IP addresses from a blog and my directory site (out of thousands of BLs)
> No inbound links from high PR sites at inception, then slowly add them
Started with links from 3s and 4s, got a few 5s, then a month later a couple 6s
> Get as many higher PageRank pages to link to your target pages as possible
Have lots (25+ PR5 and above)
> Don’t have similarities to a “directory”
Not similiar to a directory at all, very content-heavy, all unique, all written in-house

The things I have done poorly or haven't tried include:
- IBLs to deep pages (under 5% to deep pages)
- Subdomain (didn't try it)
- Outbound links to "authority" sites (haven't done much of it, just a little)
- Develop a site in a language other than English (haven't tried)

Some other ideas I have tried:
- DMOZ - listed in 2 categories!
- No duplicate content - I check every day (have GG alerts on too)
- Links from major hubs - My site is the only one that is linked to by ALL of the major hubs for my keyword phrases including 4 that rank on the first page of results @ Google

Good luck to all you sandboxed webmasters out there.

[Nacho]

Careful guys! You can try to bake the same cake with the right ingredients and get it all wrong if the recipe is not defined. I learned this from my wife's wonderful cooking.

My recommendation is: ceteris paribus

[AussieWebmaster]

Originally Posted by Nacho

Careful guys! You can try to bake the same cake with the right ingredients and get it all wrong if the recipe is not defined. I learned this from my wife's wonderful cooking.

My recommendation is: ceteris paribus

I like that latin: ceteris paribus... sadly yet realistically, very little is able to be measured all things being the same in this industry!

[orion]

Hey Nacho,

The site

Punta Ixtapa - Deluxe Beach Condo on the Coast of Ixtapa

... Punta Ixtapa is an exclusive beach resort on the Pacific Coast of Ixtapa Zihuatanejo Mexico. ... Punta Ixtapa Zihuatanejo Mexico - Deluxe Beach Villa. ...
www.punta-ixtapa.com/ - 12k - Jan 31, 2005 - Cached - Similar pages

is few days old and is ranking #15 in Google for punta ixtapa


and combined for a bit weakly recombined terms as
punta ixtapa deluxe it ranks a bit better.

Note that this site is

few days old
really has not much in-bound links
is not popular at all
it was not seo-optimized

Still it ranks high, hummmm....

This is just an experiment under non controlled conditions.

We are doing more testing under controlled conditions, which can go either way. As a scientist I must be open to test both sides of the Sandbox issue and in multiple conditions.

Orion

[Nacho]

Originally Posted by AussieWebmaster

I like that latin: ceteris paribus... sadly yet realistically, very little is able to be measured all things being the same in this industry!

Not "all things being the same" Amigo, you must leave one variable for testing and the others the same. More explicit, ceteris paribus is better defined with the following:

"A ceteris paribus assumption is often essential in all predictive sciences — in order to formulate scientific laws it is usually necessary to rule out some unspecified set of relevant factors which could interfere with the effect of some causal factor. Experimentally, the ceteris paribus assumption is realized when a scientist controls for all of the independent variables other than the one under study, so that the effect of a single independent variable on the dependent variable can be separated out. By holding all the other relevant factors constant, a scientist is able to focus on the unique effects of a given factor in a complex causal situation."

[randfish]

Orion,

I don't claim to be an expert on the sandbox, but if you are examining it from a scientific perspective, please do exmaine some keyword phrases that are considerably more difficult too! I note that there is not much competition for punta ixtapa at all:

Keyword Phrase: Punta Ixtapa
Times Searched Last Month (Overture): 0
# of Results for Search @ Google in "Quotes": 1440
# of results for search allintitle:keyword phrase @ Google: 20
# of results for search intitle:term1 intitle:term2
inanchor:term1 inanchor:term2, etc. @ Google: 23
Top 3 Bids @ Overture: $0.10, $0.00, $0.00
Strength of Competitors' Site's Backlinks: 100.00
Strength of Competitors' Internal Backlinks: 7.87/10
Strength of Competitors' Page's Backlinks: 0.00
Strength of Competitors' Pages PR: 2.7/10
Strength of Competitors' Site's PR: 6.4/10
Strength of Competitors' Size: 100.00
Percentage of TLDs in top 10 Results: 0

I have faith that you will, though. You've never let us down before. I don't really know what it is that the SEO community has done to help you out so much - you've given so much to all of us and we didn't even help you carry your sofa into your new place!

(p.s. good to have you around again)

[Everyman]

Here's a technique you missed:

Study the extent to which nonprofit, noncommercial pages and sites have been sandboxed (dot-edu, dot-org, dot-gov) by searching for keywords that would tend to exclude ecommerce. For example, "breast cancer research" might work.

Then when you've compiled your evidence, send a journalist to Scroogle to see for himself.

Explain to the journalist that Google won't tell anyone what's happening with their index, and that this has been going on for over a year. But unlike a year ago, now it's affecting public-interest sites and maybe it's time for Google to explain themselves.

[AussieWebmaster]

Originally Posted by Everyman

Here's a technique you missed:

Study the extent to which nonprofit, noncommercial pages and sites have been sandboxed (dot-edu, dot-org, dot-gov) by searching for keywords that would tend to exclude ecommerce. For example, "breast cancer research" might work.

Then when you've compiled your evidence, send a journalist to Scroogle to see for himself.

Explain to the journalist that Google won't tell anyone what's happening with their index, and that this has been going on for over a year. But unlike a year ago, now it's affecting public-interest sites and maybe it's time for Google to explain themselves.

Interesting approach....

[Nacho]

Indeed very interesting. I'm curious about the experiment however, because the terms -asdf -asdf -asdf -asdf -asdf -asdf -asdf -asdf -asdf -asdf -asdf -asdf -asdf -asdf -asdf are not part of web semantics and therefore not counted in keyword repositories associated with actual content. Eventhough that doing a search using exact match you can see that 403 results actually do exist.

Orion, what do you think? This is your specialty.

[Marcia]

Using that, there is an older site that turns up #1 for a search that was #1 for that phrase prior to the Florida update and is now not in the top 100, nor has it been post-Florida. It is too old a domain to be "sandboxed" yet exhibits the same symptoms.

[randfish]

In my opinion the use of the -asdf parameters at the end of a search string is a way to get different results, but it is not just the removal of this 'sandbox' filter. Note that results don't look like Yahoo!, MSN, Teoma or other search engines with this parameter, they look like Google has removed most or ALL of their quality filters. I don't really like the use of the -asdf at all as a method to show 'un-sandboxed', I think it's not an accurate representation of the SERPs as they would be without this filter.

[Robert_Charlton]

Originally Posted by siteseo

Create a private link network

This will put you in deeper. I'm sure that fighting private networks is one of the reasons for the sandbox in the first place.

Originally Posted by randfish

...there is not much competition for punta ixtapa at all....

randfish - You took the words right out of my mouth. From what I've observed and heard, sandboxed sites can indeed rank for phrases that are uncompetitive enough to rank on the basis of onpage optimization only.

Originally Posted by Marcia

It is too old a domain to be "sandboxed" yet exhibits the same symptoms.

Marcia - I know you know that the negative nonsense word trick worked for a while after Florida until it was disabled. It's been educational to look at sites I've been monitoring and to see that a variety of sites that I know have been nuked for various reasons (eg, closely related in several cases, or new domain name in another) come to the top with the nonsense words added. It suggests the range of what the "sandbox" might involve.

Incidentally, on a search that's been discussed frequently as a possible sandbox illustration, sma-na.org ranks for the search [sma-na].

Originally Posted by randfish

I think it's not an accurate representation of the SERPs as they would be without this filter.

You're right about dropping the quality filters, but the view with the nonsense words is exactly how the serps I watch looked a few weeks ago for about 24 hours (starting late night Jan 7), while Google was obviously running some tests.

A lot of junk also rose to the surface, and in my opinion this is why the "sandbox" is continuing... because the quality filters aren't yet working. If you believe that your site is a good quality site and that the "junk" doesn't belong there, then this could be showing you where you might end up. Or, maybe not....

[Nacho]

Originally Posted by randfish

there is not much competition for punta ixtapa at all

True in terms of number of searches using the Overture tool and also true if you compare it with a keyword that competes against millions of pages indexed. However, all keywords have their own universe of pages that they have to compete against no matter if it's 244K or 2.4 million. Still a #14 ranking for a domain that's less than a week old is not bad at all. The website went through the same process to be found, indexed and ranking algorithm. As far as I know, all sites go through the same process.

Perhaps an interesting case study on a more competitive index for a keyword and a proven achievement of how a site can get out FAST of the sandbox is electoral-vote.com. Did anyone track their actions?

[I, Brian]

A quick pointer: sandboxing isn't based on on-page factors - it's all about anchor text.

You can do fine for on-page keywords. It's anchor text pointing to a site that specifically seems affected by sandboxing.

[kservik]

Originally Posted by I, Brian

A quick pointer: sandboxing isn't based on on-page factors - it's all about anchor text.

You can do fine for on-page keywords. It's anchor text pointing to a site that specifically seems affected by sandboxing.

And the penalty that follows.

[Nacho]

Originally Posted by I, Brian

A quick pointer: sandboxing isn't based on on-page factors - it's all about anchor text.

You can do fine for on-page keywords. It's anchor text pointing to a site that specifically seems affected by sandboxing.

So, what are your thoughts for Randfish who says he's done that (see #3)?

[siteseo]

Originally Posted by Nacho

Perhaps an interesting case study on a more competitive index for a keyword and a proven achievement of how a site can get out FAST of the sandbox is electoral-vote.com. Did anyone track their actions?

I didn't, but on post-analysis it looks like typical on-page SEO...with lots of good IBL's from blogs such as johnkerry.com, harvard.edu, etc. (electoral "college" - link from a .edu...hmmmmm :-)

As far as the -exclusion filter results, I've found that many of my sandboxed sites come up #1 if I repeat a one-word competitive keyword 15 times - the same as if I use 15 exclusions. This would seem to indicate an emphasis on on-page optimization.

[Everyman]

As far as the -exclusion filter results, I've found that many of my sandboxed sites come up #1 if I repeat a one-word competitive keyword 15 times - the same as if I use 15 exclusions. This would seem to indicate an emphasis on on-page optimization.

I have a different theory. Google has more than one index due to a capacity problem. There's a meta-controller that looks at the search terms coming in and assigns it to index1, or index2, or some combination of indexes. Very complex terms are shunted off to index2 without further ado, because this is a smaller index that requires much less overhead when processing a complex search.

Multiple word searches are CPU-intensive, and 15 of them is very CPU-intensive. The meta-controller sees 15 terms separated by 15 spaces, and says, "Oh crap, this is a hairy one. Let's give it to the sandboxed index."