Sandbox, Google algorithm, Florida and MSN

[DannyS]

For the last few months I have been looking back at what happened to some of our sites in the Florida and Austin updates while looking at some new sites we launched over the last six months or so. I do suspect that what is now referred to as the sandbox started back then.

We were hit quite badly by the Florida algorithm update with a number of sites which were fairly new and were doing very well in positions on major keywords suddenly disappearing but still having very good ‘allins’. At the time all the talk was about Hilltop but these sites gradually crept back up with a linking campaign. (I have to say that one site which we left alone never did recover.) With Austin we lost some secondary keyword positions.

What was obvious with both Florida and Austin was that certain popular keyword phrases were treated differently from less popular ones. Austin just seemed to expand the number of phrases affected. The ‘allins’ no longer really tied in with the normal searches. Both of these do tie in with CBP’s list of sandbox factors. Just say for arguments sake that the major part of the algorithm updates at the end of 03 and beginning of 04 was to bring in a ‘youth dampening’ factor based on when a site entered the Google index. Wouldn’t this explain the sandbox?

In March and April last year we were affected by Google algorithm updates or additions but this for us was mainly to do with duplicate content. If the sandbox had come in at the end of 03 I guess a lot of people would be feeling it about that time period.

The big question for SEO companies would be the steepness and shape of the filter curve. How quickly does a new site start moving in major keywords and how slowly does an established site move downwards through lack of SEO work. If Google set out to hurt the growing SEO industry a factor based on site age would be the easiest way to do it. Working out this curve and ways around the filter would be the most important job we could do although I suspect a real work around doesn’t exist. Maybe just keeping a bunch of URLs alive for future needs is the only thing we can do.

And so to the MSN search. I think we could expect Microsoft to have reverse engineered the Google algorithm but the one thing they would have to think twice about would be basing results on age in the index. Their original index has been expanded so much that it would take out the majority of the sites in competition for positions. For us the Google ‘allins’ seem to tie pretty closely to the MSN raw results.

Danny

[cspelts]

Maybe just keeping a bunch of URLs alive for future needs is the only thing we can do.

That's not something I had thought of before. Figuring out future needs is the tough part, but it might be worth guessing. Thanks for giving me something to consider!

[tcampione]

There is debate as to whether sites get sandboxed, or links do. I tend to lean toward the ‘sandboxed link’ theory, primarily because I’ve heard of some older sites (although not many) experiencing something similar to the sandbox effect.

Through trial and error (mostly error) I’ve come to believe that the speed at which BLs are added to new sites can keep a new site’s BLs sandboxed. Adding BLs in a slow steady pace rather than in bunches is probably a better tactic. It is of course quite easy to track the percent increase of BLs in a given period (three weeks for example) and just as easy to penalize for a large increase.

I think it is also very possible that the maximum number of BLs for each site over its lifetime (perhaps) might also be a part of the site’s record. So, if an older site had 1000 BLs two years ago, only 600 now, and were to add 300 within the next few weeks it would probably not suffer. But if it added 500, I think there is a great chance that would.

{All of this is pure speculation based on minimal observation.}

[jacobwissler]

I have noticed the result of good anchor text remains long after the link is gone. If a major website links to you for one week, the anchor text appears to remain in the search engine memory, sometimes for years. Therefore, the number of current links is not the only factor; the history of the site, and past links, will appear to have an impact.

[stephenmunday]

What I don't understand is that if the "sandbox" or "sandbox" link theories are true, then Google is:

1) Penalizing new sites
2) Penalizing sites that gain sudden popularity.

But doesn't this contradict one of the main points about the Internet? When I search, I want the most current, relevant information.

Say I set up a political site like JibJab. It goes live, I get some good publicity through people picking up my PRWeb press release and I start getting a bunch of visitors and a lot of unasked for IBLs. Both the "sandbox" and "sandbox link" theories would have Google penalizing my site for months. It may be wildly successful and written up in all the main print media, but Google would be suspicious of it and ignore it.

Why would Google want to do that? Wouldn' Google users want to find this great site?

Yes, Google want to filter spam, but why would they want to sacrifice one of the main indicators of quality results - timeliness - to do that?

This is why I am wondering whether there might be some other explanation for these phenomena. Could there be any other reason for what we are seeing?

[Like2Golf]

Originally Posted by DannyS

The big question for SEO companies would be the steepness and shape of the filter curve. How quickly does a new site start moving in major keywords and how slowly does an established site move downwards through lack of SEO work. If Google set out to hurt the growing SEO industry a factor based on site age would be the easiest way to do it.

I totally agree. The top listed sites for my keywords have been around for a long time, and they even SPAM (I've reported them, but nothing's happened) yet they remain on top.

My site, which I have been very agressive at optimizing, is ranking in the 90's.

On MSN - I'm #1.

[Intensity]

Good point. There is pretty nice discussion going on about something very similar on the SEOCHAT forum.

In the discussion, a theory has been made about a filter that keeps new sites Filtered, not sandboxed, to prevent a sudden surge in the SERP's.

A good theory has been made that possible the age of the backlinks plays a factor in the results of your SERP's.

Check it out!

[freehits]

looks like I just need a product named
"-sqfqsdfqsd -sqfqsdfqsd " and I could never worry about sandboxes again.

[Intensity]

LOL. You got that right.

[bbott]

Originally Posted by Intensity

In the discussion, a theory has been made about a filter that keeps new sites Filtered, not sandboxed, to prevent a sudden surge in the SERP's.

Whether it's a filter or sandbox, what is the real difference? Either way your site is being withheld from ranking as it would w/o any filters, etc.

For instance, my website has been around for 10 months and we're still in the sandbox w/ Google. We rank top 10 for numerous keywords on MSN, Yahoo!, Altavista, etc. - no where to be found on Google.

When we try the sandbox tool on the SEOCHAT forums - we score an 80% which is a very good indicator we're being withheld in some way or another.

Also, the "-sqfqsdfqsd -sqfqsdfqsd " technique shows our site in the top10 for numerous google keywords.

To be quite honest - I'm getting sick and tired of google! Yahoo! and MSN have my full attention.

Sorry if I rambled off topic! :)

[stephenmunday]

Arrrrrgh!

I'm top for my main keywords with the sandbox effect taken away! Thanks for the link to the seochat post, guys.

Could you also post the link for the sandbox tool you are referring to. I must be dense, but I couldn't find it.

[incrediblehelp]

Originally Posted by Like2Golf

I totally agree. The top listed sites for my keywords have been around for a long time, and they even SPAM (I've reported them, but nothing's happened) yet they remain on top.

My site, which I have been very agressive at optimizing, is ranking in the 90's.

On MSN - I'm #1.

I disagree. Many times when I work with clients they have been online for sometime, they just need guidance on organizing their website so that the search engines can properly spider their content. I don't see why Google would sandbox a new website. Their are to many reasons not to demote it. I do believe the theory of sandboxed links discussed above, when adding many links at once, which could start the sandbox gears turning.

[DannyS]

Accepting that there is a sandbox, is it the age of the site being indexed or the age of the backlinks that kicks it in? The former seemed a good theory as it would be so easy for Google to do. All it would need would be to filter on small date field in their index. If it were on the age of backlinks that would destroy my idea of keeping some URLs live for future sites;-(

Why would they have a sandbox is interesting. If they set out to hurt SEO this would be such an easy way of doing it. What would a SEO company be able to promise a client in Google results? “Wait a year and you will start to show in your major keywords.”? We can guess that Google doesn’t really like SEO work as they would feel it could distort their results. Hilltop had nothing in it aimed at SEO but the sandbox, duplicate content filter and long waits for a PR update do. The latter was a good way of slowing the sale and buying of text links based on high PR. I suspect that for the engineers at Google their number one enemy isn’t MSN or Yahoo but is the SEO industry.

Using nonsense phrases in keyword searches, e.g. the –sasdf repeated 16 times or whatever can show unfiltered results, but maybe it’s more than just the sandbox filter that gets turned off. I know it makes little difference now, but does anyone have information on the age of sites or linking campaigns on sites that were hit badly by the Florida update. I still suspect that the sandbox started back then. The problem is that probably by changing one digit in the algorithm the affect of a sandbox filter could be changed significantly and Google could be fine-tuning this often.

The only post I’ve read so far on getting out of this sandbox is by nuking the site with very high PR paid for text backlinks. Has any one seen or heard different?

Danny

[incrediblehelp]

Originally Posted by DannyS

Accepting that there is a sandbox, is it the age of the site being indexed or the age of the backlinks that kicks it in? The former seemed a good theory as it would be so easy for Google to do. All it would need would be to filter on small date field in their index. If it were on the age of backlinks that would destroy my idea of keeping some URLs live for future sites;-(

I truly believe it is a little bit of both. More on the IBL side for the sandbox theory.

Originally Posted by DannyS

Why would they have a sandbox is interesting. If they set out to hurt SEO this would be such an easy way of doing it. What would a SEO company be able to promise a client in Google results? “Wait a year and you will start to show in your major keywords.”?
Danny

I try to sign clients for 12 month campaigns.

[Daniel Brandt]

There is now a Scroogle2 demo of the new "Florida2 filter." This time lots of dot-orgs and dot-edus and even dot-govs got hit. Try some variations having to do with cancer research, and you'll see what I mean. This filter uses the string of -adsf (about 15 of them) to look at what the top 100 results are, with and without this string.

Whether it's a filter or not is anybody's guess, but it's happening in real time, and if your site got hit I can imagine it's rather annoying regardless of what you want to call it.