To clarify what this is about, the official definition of the sandbox effect is: “when your website carries a high position in Yahoo and MSN, but is nowhere to be found in Google”.

Many people will maintain that the sandbox does not exist. I agree that there is no sandbox “filter”, but there certainly is a sandbox “effect”. We cannot know its mechanics in detail so theoretical discussion on the matter has little value. I prefer to concentrate on results.

Here’s another one of my experiment blogs. It will give some hope to people who have created new sites and put countless hours into backlink resourcing, only to see their site vanish from the google serps.

I registered a new domain in early March 2009, and it was indexed soon after. I wanted to get 20 subpages ranking for multiple competitive marketable keywords.

Note: No linkwheels have been applied to this site

What I did:

* April – Start a major linkbuilding campaign. I was working under the simple mantra: more high PR dofollow comment spam = higher rankings. No attention was given to relevance or any other factor. Therefore, a great many of the links were extremely spammy, some on sites with thousands of outbound links.

PRE-RANKING PERIOD

* June 20th – By this point I had 300 backlinks to home page, 100 backlinks to subpages 1-12, 300 backlinks to subpages 13-18, and 1000 backlinks to subpages 19-20. Prior to this date, none of my pages ranked anywhere. After June 20th, however, I began to see my pages in the top 5 in most cases on Google, Yahoo and MSN.

* June 22nd – Removed from Google serps.

* June 24th – Return to June 20th Google position.

* June 26th – Removed from Google serps. As of June 26th, all backlinking ceased immediately

SANDBOX – 2 MONTHS

* September 1st – Begin to see all pages return to the Google serps, between page 3 and 10.

* September 15th – All 20 subpages return to page 1. 6 subpages take position 1, 14 subpages land in the top 5.

I can’t believe how well this has worked out. Since June, I was working under the assumption that I had done something wrong with my backlinking, and turned to a sort of backlink purism. I conducted another experiment on this here, with a different domain:

This is a classic open and shut case of the sandbox effect. It is characterized by an alternation between domination and total absence from the serps. Something more than the “google dance” is going on here.

It is often conjectured that the sandbox is a test to see if links continue to come in after serp removal. My experience speaks against this theory.

One final note: Google has ranked one of my pages number 1 for a highly competitive software product. The page has over 1000 spam links with same or similar anchor text. None of these links are from relevant sites. Perhaps we should be less paranoid about google’s ability to detect us.