Sep
27
Google Sandbox and Usage Statistics
Filed Under Google | Comments Off
Aside from a number of things that new sites can be filtered for, causing what appears to result in the “Sandbox” effect or “Ranking Delay” which causes newer sites not to rank for their keywords for a long time, with inbound links being one of the suspected factors, isn’t it possible that the initial traffic level the site generates can be a factor that determines whether or not it hits the Sandbox?
Assuming it’s nothing more than a combination of several (or many) filters why wouldn’t measuring traffic be a consideration? It’s been used since back in Direct Hit days, and while in itself isn’t enough basis for ranking sites, it can’t be denied that apparently search engines considered it of value else they wouldn’t have bothered to use the measurements.
There was a patent issued to Google about using traffic and usage statistics in scoring site back in 2002: Methods and apparatus for employing usage statistics in document retrieval That could possibly be related to why some people think Adwords helps avoid the Sandbox. While not Adwords traffic or participation itself, it could be a combination of factors, that combined with different sources of exposure, bring in a lot of inbound visitors from the start for some sites – and not others. And then, if there are multiple pageviews per visitor and/or a decent amount of time then spent on the site per visitor, so much the better.
There have been members at forums who reported no “sandbox” for their new sites who may not have gotten media or other high profile exposure or links, as some recommend doing, but they had linked to their new sites from their older, established, high traffic sites which sent a considerably heavy volume of traffic to their new ones – no sandbox.
I can’t shake the idea, after all this time since first reading that patent, that “usage statistics” might be a somewhat significant factor, and the fact that new sites have temporarily appeared for searches in SERPs and then disappeared could give suspicion that clickthroughs and bounce rate were being measured as a metric toward building the initial trust and desirability factors for a site, which could help indicate whether or not it’s a site that’s relevant and to be trusted.
On another front, it could also be theorized that traffic and usage stats are something that can contribute to what appears to be a penalty on affiliate sites. There may not be one as such, (though there are factors with some that could bring one on unrelated to their being affiliate sites) but a visitor coming to an affiliate site (and they don’t generally get high profile exposure or a heavy initial traffic influx), may stay for a few minutes, if that long, and then click off to a merchant site, where they’ll stay, browse and maybe hit the shopping cart.
Given that search engines dating as far back in my own personal memory of search as MSN using a combination of Looksmart and the Inktomi database with Direct Hit factored in, and the Yahoo Directory back then being ordered and used for traffic being heavily influenced by traffic/usage to the point where some out there were running clickbots to up their rankings, IMHO it would be naive to think that usage factors aren’t currently playing at least an appreciable role.
After all that’s said and done, maybe those who advocate getting (or buying) links for the traffic rather than for PageRank aren’t just preaching pedantic purism. And maybe people who advocate using the nofollow attribute aren’t giving such bad advice as some would like to think. If nothing else, even if the usage statistics theorizing is full of holes, getting traffic even with nofollow is NOT a bad thing until (or even after) rankings start to roll in. Anything that brings targeted traffic is just plain good marketing, though sometimes the overall picture of marketing can get a little blind-sided with the seo-minded.