The Sandbox (a.k.a. Sandboxing or the sandbox effect or the Google penalty) is a name given to an observation about the way Google ranks web pages in its index. It is the subject of much debate—its existence has been written about[1][2] since 2004[3] but not confirmed, with several statements to the contrary.[4]
According to the theory of the sandbox effect, links which may normally be weighted by Google's ranking algorithm, not least improving the position of a webpage in Google's index, may be subjected to filtering to prevent their having a full impact. Some observations have suggested that two important factors for causing this filter to come into play are the active age of a domain, and the competitiveness of the keywords used in links.
Active age of a domain should not be confused with the date of registration on a domain's WHOIS record, but instead refers to the time when Google first indexed pages on the domain. Keyword competitiveness refers to the search frequency of a word on Google search, with observation suggesting that the higher the search frequency of a word, the increasing likelihood that the sandbox filter effect will come into play.
While the presence of the Google Sandbox has been long debated, Google has made no abject disclosure on the matter. However, as the sandbox effect almost certainly refers to a set of filters in play for anti-spam purposes, it is unlikely Google would ever provide details on the matter.
Google has long been aware that its historical use of links as a "vote" for ranking web documents can be subject to manipulation, and stated such in its original IPO documentation. Over the years Google has filed a number of patents that seek to qualify or minimise the impact of such manipulation, which Google terms as "link spam".
Link spam is primarily driven by search engine optimizers who attempt to manipulate Google's page ranking by creating lots of inbound links to a new web site from other web sites that they own. Some SEO experts also claim that the sandbox only applies to highly competitive or broad keyword phrases, and can be counteracted by targeting narrow, or so-called long-tail phrases.[5]
A "reverse sandbox" effect is also claimed to exist, whereby new pages with good content, but without inbound links, are temporarily increased in rank—much like the "New Releases" in a book store are displayed more prominently—to encourage organic building of the World Wide Web.[4][6]
David George disputes the claim that Google applies sandboxing to all new web sites, saying that the claim "doesn't seem to be borne out by experience". He states that he created a new web site in October 2004 and had it ranked in the top 20 Google results for a target keyword within one month. He asserts that "no one knows for sure if the Google sandbox exists", and comments that it "seems to fit the observations and experiments of many search engine optimizers". He theorizes that the sandbox "has introduced some hysteresis into the system in order to restore a bit of sanity to Google's results".[4]
In an interview with the Search Engine Roundtable website, Matt Cutts is reported to have said that there are some things in the algorithm that may be perceived as a sandbox that do not apply to all industries.[7] Jaimie Sirovich and Cristian Darie state that they believe that, while Google does not actually have an explicit "sandbox", the effect itself (however caused) is real.[6]