Matt Cutts on the Google Sandbox

51 comments

A thread at WMW starts with a pubcon snippet, that is hardly earth shattering, but is none the less worth recording.

Quote:
Matt said that there wasn't a sandbox, but the algorithm might affect some sites, under some circumstances, in a way that a webmaster would perceive as being sandboxed.

and a further addition in the thread is:-

Quote:
I got the impression that it only affects sites in "certain industries," perhaps from some of Matt's subsequent remarks.

and

Quote:
I was at the Q&A and listened to Matt's response. The part that I thought was interesting was that Matt said when they (Google) first started hearing about the "sandbox" as the term is used by webmasters they had to look at their algo to see what was causing it and then look at the sites it was affecting. Once they studied it, they decided they liked what it was doing.

So what I concluded was:
1. Yes, there is something that behaves like a "sandbox."
2. It's part of the algo, but wasn't necessarily created to do what it does. (they didn't set out to build a "sandbox.")
3. It affects some site and not others.
4. Google likes what it does.

So, take out any mention of "hotels" then.

Comments

certain industries

I got the impression that it only affects sites in "certain industries,"...

Yes, only industries with web sites :)

It's a SPAMBOX!

If Matt's right and it's anti-spam algos at work then call the thing a SPAMBOX so we know we're on the same page :)

Plus, as I've said many times in many forums before, some people just can't get past the fact that the 10,000 web sites fighting over their keyword already have a link network firmly anchoring them in the SERPs.

Seriously, what kind of narcissist thinks that just because he can't break into the top 100 on a keyword means there's something keeping you out other than the 100 sites that came before you?

Did you ever consider that they were there first and better anchored?

NO CUTTING IN LINE BUDDY!

Sheesh, bunch of whiners, oh boo hoo, sandbox...

it doesn't work

Whether Matt or anyone else at Google likes what this part of the algo does is beside the point. It doesn't work. Not only should any webmaster take the word "hotel" out of their website, anyone actually looking for info on travel in general would waste time searching Google cause there's so much crap that comes up in the search.

they liked what it was doing = bias

A classic example of sample bias.

If only competitive webmasters know enough to report evidence of sandboxing (or even discuss it), then naturally the Google engineers (little "e") will like what they see. All they see are competitive websites getting bridled.

VINDICATION!!

I have been saying for forever that there is NO Sandbox, I am soooo glad Matt finally vindicated me.

As far as I am concerned the only ‘news’ is that google did not intentionally set out to create this particular ‘filter’ module in the algo, that they stumbled upon it by accident.

All this ‘sandbox’ is, is just a LOT of people that don’t understand how google ranks and filters pages/sites, so when one of these peoples sites tanks or does not preform as well as expected, it is very easy to call it the ‘Sandbox’.

A free bit of advice - For those of you that believe you are suffering from the so called ‘Sandbox’, take a good hard look at your internal linking structure...

Hmmmm

Yeah... SEOmoz couldn't rank for its own name (which never existed on the Internet prior to my creating it) for 12 months because that's just how Google ranks sites...

lots0 - I may not agree that sandbox is a good name for the effect and I may not agree with many folks who give the effect attributes it doesn't carry. But, to say that it's just poor internal linking structure or just the way Google ranks sites seems even more naive to me.

Otherwise, why would a site that has no changes to it suddenly go from ranking 35+ for its own name to #1 across the board for everything in any of its pages' title tags? That's why "escaping the box" is such a great descriptive phrase...

..

But, to say that it's just poor internal linking structure or just the way Google ranks sites seems even more naive to me.

I gave a 'free' piece of advice, you are free to take it or leave it.

If you want more detailed information, like how not to trip the 'sandbox' filter your gonna have pay me and sign a non-disclosure agreement... Now how naive is that?

>>>added

why would a site that has no changes to it suddenly go from ranking 35+ for its own name to #1 across the board for everything in any of its pages' title tags?

I can think of several reasons for this that do not involve the 'sandbox'.
A) you acquired a 'good' link.
B) a change in the algo.
C) Without realizing it, you fixed the problem.
And there are more, but I am short of time right now.

Also worth to mention,

Also worth to mention, again, is that the effect of the so-called sandboxing has linguistic components to it that really limits it to a very small segment of the world. Yes, my good american friends - take a look at a map, USA is just a small area to the left :)

Linkage at all

Lots0, that's only half of the story. Internal linkage is an important factor to determine reasonable (read: what Google finds natural and user friendly) linkage, but it's seen in the context which is defined by external linkage and it's amplification/distribution within a site, also considering the attitude on outgoing links, placement of links, and so on. One must be an unselfish noob ignoring all SEO advice, or a smart SEO, to escape the dubious 'sandbox-effect'.

Baloney

Sounds like baloney to me. Then again, Google never misinforms webmasters, right? 302s only affect spam sites, you can't hurt competitors, etc.

If there is no sandbox, why do sites magically pop out one day?

Sandbox or whatever you call it, there is a filter that targets competitive terms. These competitive terms typically generate a lot of adwords revenue. Do the math.

...might affect some sites,

Quote:
...might affect some sites, under some circumstances, in a way that a webmaster would perceive as being sandboxed.

More like most sites, under most circumstances.

I'm not complaining as we have many older sites, and clients who have the same. We have had some difficulties with new sites, and a few recent successes. We are actively analyzing our now rather large list of sites that have escaped the SpamBox (ranking for reasonably competitive phrases with a wide breadth of industries... even a "+ hotel" term or two) and hope to continue drawing conclusions from these studies so that our sites may join the list. :-)

..

Mikkel, ;-)

Sebastian, I guess it boils down to how you define ‘internal linking structure’. I think my definition of ‘Internal Linking Structure’ is rather broad and covers about all aspects of linking pages within a domain.

thePhenomenal, It very well could be, but as Matt’s statement about the ‘sandbox’ agrees with what I have been saying all along, I chose to embrace his statement. :-)

Great article, Sebastian!

Great article, Sebastian! Thanks.

Sandboxing was originally

Sandboxing was originally used to describe a specific effect - that of volume link building not having the same effect it originally had.

For those who have never seen it - congrats on missing that part of the filter. But don't believe just because you don't see a thing, that it therefore does not exist.

..

But don't believe just because you don't see a thing, that it therefore does not exist.

I have studied dozens and dozens of sites that were susposed to be under the new site sandbox effect (real estate, gambling, drugs and mom and pop stores) and I have still never seen it. So yes I do find it hard to believe.

Brian, if you have an example, I'd love to see it, PM me with it.

>>>added

Sandboxing was originally used to describe a specific effect - that of volume link building not having the same effect it originally had.

My understanding was that 'sandbox' was used to describe new sites that failed to rank until some time limit was reached...

I guess if we don't have a common definition of 'sandbox' there is no sense in continuing this discussion. :-)

Example

lots0, I'm not disagreeing with your theory. But do you have an example of a site that is less than 6 months old that is ranking for a competitive keyword phrase?

That would be something to

That would be something to see. I've only ever stumbled across a few. BTW - lots0, I'm starting some new sites in 2006, so I'll be happy to pay and sign if you can keep them from getting boxed. I'll email you!

Matt clearly disagrees with lots0

He said the effect was unintentional. As reported, his comments imply that the engineers did indeed find a cause for the effect.

So, the Sandbox Effect is essentially proven to exist.

Q.E.D.

Sandbox people seem to be

Sandbox people seem to be like cat and dog people unable to reach a common ground for discussion, so these things deteriorate into a pissing match, which doesn't do anyone any good. For those of you who don't believe if you want to be helpful share some ideas to help others. For those of you who do believe think about what you are doing on sites that are out, as opposed to sites that are in, and think of ways to change them. To start the ball rolling here are some of mine:

  • Come to grips with your crappy link exchanges and dump them, keep your good link partners.
  • Is your content really as good or better than what ranks now? If not correct it.
  • Spend a full day analyzing the backlinks of people who rank for your terms. Go after the ones at the top of the list not the bottom
  • Create a tool, run a contest, give something away, give people a reason to link to you
  • Don't change too much too quickly
  • Develop things in parallel with different aproaches, compare results, see what works and what doesn't

I agree with MM

[A change is as good as a rest]

Common Sense

1. Take a rather competitive keyword (not time sensitve, like Hurricane or something like that).
2. Save the URL's in a notepad
3. Visit a whois registrar and search for when the sites were regsitered
4. Notice the registered date
5. Compare for the top 10 sites.

I don't think there's much else to say?

In English in the US of course

Agree with Mikkel

So the problem is...

Graywolf
Come to grips with your crappy link exchanges and dump them, keep your good link partners.
Is your content really as good or better than what ranks now? If not correct it...

When I see crappy sites like this, I don't really look too deeply. It's with sites (like SEOmoz), where there has never been unnatural links, it was developed 100% organically and really does have 40,000+ domains pointing to it because they liked what they saw.

When those type of sites don't come out of the box for a year, I'm more and more convinced. And it's not like SEOmoz is a competitive keyword... :) Even now that the site ranks, we still get less than 1/5 of our traffic from search engines (110-150 visitors per day from Google). So... why are they boxing? Is it really only certain industries?

Sorry - Not sure why I'm so defensive about this issue. Must be close to my heart, I suppose.

beyond SEO

concerns, the effect on search results is apparent and gives a pretty creepy effect, like they are hiding things, almost scary to think they do it and people still think google is the authority.

I don't trust their serps anymore and rely on engines that keep fresh snapshots. better info.

spam fighters gone wild.

Rand from the eyes of the

Rand from the eyes of the big G I would say many sites in the SEO community live pretty close to the wrong side of the tracks, and that's probably where most of your initial links came from. Not saying thats good, bad, right or wrong, it just is. I suspect you came out shortly after your big algo analysis, which was making the rounds on the digg front page and technorati most popular list, and probably got you a considerable amount of natural links from sites outside of the SEO link community.

To be honest I haven't analyzed your site or backlinks, but that's the way I remember it off the top of my head, please correct me if I'm wrong.

lots0, I'm not disagreeing

Quote:
lots0, I'm not disagreeing with your theory. But do you have an example of a site that is less than 6 months old that is ranking for a competitive keyword phrase?

Yes.

Quote:
That would be something to see. I've only ever stumbled across a few. BTW - lots0, I'm starting some new sites in 2006, so I'll be happy to pay and sign if you can keep them from getting boxed. I'll email you!

I doubt it's cheap, you can probably study the sites ranking and see what's working.

Build a script, go hunting, it's kind of lame to out people here who are pulling it off. There are hundreds, probably thousands of 2005 domains ranking for respectably competitive terms.

Links Schminks

My wife and I launched a few sites this year that rank quite well but they weren't really busy terms - the site on a really popular term is struggling, so I think part of the problem has to be the popularity of the niche you're targetting.

Lastly, imagine if 100 of us @ TW target the same keyword that we can't all get the top 10 - so if that makes no sense to you, then you'll have a long and strange trip.

Graywolf - you could be

Graywolf - you could be right - SEOmoz popped out about 5 weeks after the ranking factors article. But, it seems so odd that the same day, a dozen or more other sites all started ranking for lots of terms that they previously only ranked for with allinanchor searches and at Yahoo! and MSN.

That's what baffles me. Why do all the boxed sites jump out of the box together on certain days. It's almost like GG has a holding cell and then, one day, the jailer decides it's time to free another round of prisoners and voila, a few hundred sites that couldn't rank for their domain names are ranking in the top 5 for dozens of terms.

No matter what we want to call it, it's a highly observable effect that can be easily repeated. And it most certainly doesn't just happen to SEOs or their sites.

I liked this contribution to that thread...

Quote:
Question for Matt Cutts:
"Does gravity exist?"

"Most say yes. The fact is that there are certain observable physical phenomena that could be perceived as a 'gravity effect.'"

Rand

From a quick view on your link neigborhood I think Graywolf hit the nail on the head. As for time limited 'phenomenons', Google has a history on that. For example the 30 day bans for hidden links, although that's a penalty for intent, and the extended probation period is kinda caveat to evaluate intent. Probably the concurrently freed sites are a result of periodical batch processes looking at a bunch of sites of a particular state.

My understanding was that

My understanding was that 'sandbox' was used to describe new sites that failed to rank until some time limit was reached...

Indeed, the entire sandbox concept has been quite expanded, but it seems that age of domain and general "authority" processes have been a big part of it.

It is a very abused word, though - to some "sandboxing" is nothing more than a metaphor for "crap site with no real SEO".

The process is very real, though.

While I don't have the keys

While I don't have the keys to the box watching multiple sites come out and some drop back in is actually very educational. To the best of my understanding/observation there is an evaluation that is done around the same time that PR/backlinks are updated, and that's why sites come out all in once.

If you keep developing sites that are getting boxed, stop developing them that way. If you want to watch the algo at work develop two sites about the same subject and promote them with opposing methods at the exact same time.

Lastly you don't have to be about SEO topics to be part of the SEO link community

What about aging delay?

Could someone instead ask Matt if there is an aging delay?

If he says no, then it would be clear that he is full of shit! :)

Jill, your oblique reference eludes me...

What are you really trying to say?

>If he says no, then it

>If he says no, then it would be clear that he is full of shit!

No, he would be telling the truth. The sandbox is nowhere near as simplistic as you seem to want to believe. You need to clear your mind of this aging obsession and think more in terms of the old style Ink BOW, hth.

That's the point

NFFC: "The sandbox is nowhere near as simplistic as you seem to want to believe. You need to clear your mind of this aging obsession..." should be the title of this thread.

Sandbox - No Sandbox

Whats wrong with half you people.
If it quacks and waddles --- its a duck.
Call it what you like but its still a duck and always will be a duck.
Try telling my accountant that there's no sandbox.

Matt Cutts

Just one more thing. Matt Cutts works for Google and is probably a very rich man on paper. Does anyone think he would piss it up the wall by making injudicious comment. Break him in two and he probably has 'Google' printed full length and who can blame him. Anything he says must be considered from the Google corporate position. Dont read to much into it.

I believe, (like it matters

I believe, (like it matters what I believe), the reason for the confusion is that both sides of the issue are right. That makes it very difficult for either side to concede while the need for acceptance by both camps is even greater. = elevated emotional attachment to an abstract concept.

NFFC is giving everyone the most obvious clue. It's not about aging. It's about trust.

I can unequivocally state from personal experience that if you have a site you built at 1pm today and tonight at 7pm CNN, MSNBC and Jeremy Zawodny linked to it, AND if those links made up a vast majority, (if not the entirety), of your inbound links, by 10pm you would be placed very well for hundreds and maybe thousands of keywords and phrases.

If over the next few days you got another 63,000 links from auto generated, bullshit blogs and 300 more from your throw away domains that you've been spamming for the past 2 years, that same site will drop to #111 and keep dropping until finally it had it's cache removed in Google.

My point is that if your site made news and made it first, do you really think Google wants to put a site like that in any kind of box.

As far too often is the case with all things SEO, (whatever that is), we tend to fall into the trap of only seeing a picture from our prospective. Even the very question is being asked of Matt Cutts from an SEO perspective and with little regard to what Google's perspective must be.

If you can find a way to stop seeing it as YOUR problem and try to see it as GOOGLE'S problem, the sandbox explanation falls away leaving nothing but the foundation of TRUSTRANK. Not as we know that term from the paper we all discussed a couple of years ago. Thinking of Trustrank in those terms is again falling into the trap of seeing the problem as OUR problem and not Googles. Rather think of it as NFFC alluded to. BOW

If you can start seeing it that way, NOW go look at your backlinks. Then look at the backlinks of those backlinks. Do you think GOOGLE sees those as trusted or in question?

The good news is, wholesale blog spamming is coming to a long overdue demise. The bad news is, all those millions and millions of worthless, nothing but a silly auto generated feed, no human would ever want to read, blogs by the truckload are likely to be the new BH/WH doorway page circular debate for a long time to come. Anyone wish to fathom a guess of how many posts we will all have to endure claiming:

I've never done anything spammy on any of my 142 blogs

over the next couple of years?

...

thePhenomenal said;

...do you have an example of a site that is less than 6 months old that is ranking for a competitive keyword phrase?

See, this is where the problem starts, six months is NOT enough time to rank a new site for any competitive keywords in google. Even assuming that you are using aggressive SEO, six months is just not enough time to rank for competitive keywords.

I don’t hear a lot of people saying that their niche sites in non-competitive areas are having problems ranking in the first few months...

It all goes to competition.

The reason that new site is not ranking for the keywords ‘gambling/real estate/porn’ after 3 or 4 months, is rather obvious at least to me.

There are thousands and thousands of other sites competing for the same keyword, some very well financed and with some very smart SEOs working for them.

So it is only common sense that it is going to take a new site/page a lot of work and a lot of time to even rank in the top one thousand results for any real competitive keyword, they are playing catch-up with the established sites/pages and the established sites have a huge (maybe even an unsurmountable) head start.

People seem to forget that thousands of other people are trying for that compeditive #1 spot. The sites/pages that are already there want to stay there so they work at keeping their ranking everyday and everyday there are also more and more people building more and more sites/pages targeting that same keyword.... and google only lists one thousand results for any keyword.

For those of you that pfffft me, it won't be the first time I have been pffft'ed. I was pfft'ed when I first stated in public that google wanted to move to a constant indexing system (some arsewipe told me that the monthly google updates were a 'keystone' in how google works and they would never get rid of them...) pffft...

I was pfft'ed when I was one of the first to state keyword anchors carried more weight in the algo than keyword text.

I was pffft'ed (It was more intense than that, some people wanted to burn me at the stake) when I was one of the (if not the first) first to say in public that links were the key to ranking.

Time will tell if I am right or wrong, I can wait. ;-)

Pfffft

Quote:
...I was pfft'ed when I first stated in public that google wanted to move to a constant indexing system (some arsewipe told me that the monthly google updates were a 'keystone' in how google works and they would never get rid of them...) pffft...

I was pfft'ed when I was one of the first to state keyword anchors carried more weight in the algo than keyword text.

I was pffft'ed (It was more intense than that, some people wanted to burn me at the stake) when I was one of the (if not the first) first to say in public that links were the key to ranking.

1 out of 3 is not the most impressive track record, though, so I'll continue to accept what Matt said about the Sandbox Effect being an unintended effect that was allowed to continue as somewhat more reliable than your (and other people's) denial of any such effect.

Massa, I do believe you are

Massa, I do believe you are exactly right. And Lotso, you're exactly wrong. With NFFC somewhere in the middle!

Jill

I don't disagree with NFFC or the other guy, so how could I be exactly wrong? Rhetorical Question... :-)

Because with this

Because with this comment:

Quote:
So it is only common sense that it is going to take a new site/page a lot of work and a lot of time to even rank in the top one thousand results for any real competitive keyword, they are playing catch-up with the established sites/pages and the established sites have a huge (maybe even an unsurmountable) head start.

You've shown that you haven't seen what many others have seen when it comes to new domains.

It's not at all the same thing as the usual taking time to see results thing that we are all familiar with from years in SEO. This is completely different. You literally cannot be found for exact phrases in title tags that are not competitive at all.

Then one day you can. And you can be found for most of the real phrases you optimized for. It's like night and day.

There is a "retardant effect" but no "aging delay"

There is a "retardant effect", but it's not an aging delay. It lasts as long as Google needs to evaluate its quality indicators, or the lack of them.

..

Jill said,

You've shown that you haven't seen what many others have seen when it comes to new domains.

Jill, if you would have read what I wrote in this thread you would have seen where I wrote that I have never seen this so called sandbox, even after looking for it.

MM, I am sure that Matt will be jumping with joy and his eyes filled with tears of happiness to hear that you feel that way.
I know I am really impressed with you.... not.

BTW - It is four out of four...AFAIK

NFFC, it's like that Spam

NFFC, it's like that Spam Mass patent by Yahoo?

Regardless of what they call it

The general effect is an aging delay or sandbox.

Matt specificially said (paraphrased here) "We have no sandbox but it can look that way though"

Thats all I care about. He affirmed my model of the search engine is correct up to a point. That model has a sandbox in it brought on by a delayed utilization of incoming links to determine ranking factor.

or delayed verification

>That model has a sandbox in it brought on by a delayed utilization of incoming links to determine ranking factor

I dont think its as cut and dry as just "delayed utilization" of links. I think one of the reasons some people have sites that appear to avoid the sandbox is because of the type of link or authority factor of the incoming link.

Performancing.com will serve as an intersting test to that theory, as it is getting some damn good authority links. Too bad we all cant rank as fast as Matt Cutt's blog, though.

waves to 2much

>Spam Mass patent by Yahoo

Yes, that type of stuff shows the way the thought processes are going. I think the main change has been more of a fundamental shift, they seem to have moved from the "everybody is good until they prove otherwise" to "everybody is scum until they prove otherwise" mode. Time is one factor in that, prolly the biggest factor but not the only one.

Up to them what they do with their engines but I don't think its a good move long term. It has the effect of making "tricks" a requirement to rank at least with new sites, opens the incentive for web "spamming" to a much wider audience. As I have alway said its not the professional spammers they should be worried about, its the amateurs that do the damage.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.