|Has the "sandbox" slackened off? How about new pages, old domains?|
We've noticed something odd the past couple weeks, at least for us. Google continues to find and index our new pages very quickly (we aren't a big site, may add a few dozen pages per year), and in the past, we'd always see some traffic as soon as pages were indexed.
Now it seems our new pages are being treated like they are in a sandbox. They appear in results for super specific searches that nobody would make, so Google has indexed them, but they aren't getting any natural traffic at all. Zero. It's not a question of relevence, the content is a good fit for our site.
Anybody else seeing Google treating new pages on a long established website as if they were on a new website, rather than an old one?
The last three websites I have launched have all been indexed and ranked very quickly. One of them in particular was attracting search traffic two or three days after launch with only one IB link from one of my own websites.
I launched a small site for a client last week. He hires out Green Widgets. He emailed me yesterday to tell me he was in position 5 for green widget hire. Clearly he was chuffed at this but I warned him about the new site boost.
Having said that, the other two sites I launched also received traffic within a few says of launch and it dd not drop off the way it used to when new sites were classically "sandboxed".
I have been very busy lately. Have I missed something or have I just got lucky?
[edited by: tedster at 8:23 pm (utc) on Sep. 18, 2007]
[edit reason] moved from another location [/edit]
My recent experience is that sandbox has become more "subtile" than it used to be.
The term "sandbox" might not be the most apropriate. I still see signs of "over optimization penalties" for some new websites, but only on very competitive keywords and only if there is over optmization.
Now if you target long-tail keywords and don't go too hard on backlinks you can start ranking in Google SERPs few hours after the site is published, and there is no big drop to expect on your keywords.
It looks like since you start ranking with only one IBL, you are in this case. You should be fine as long as you continue growing your backlinks smoothly.
Our experience over the past 6 months has been a dramatic easing of the sandbox effect. New sites can rank rather quickly with secondary terms ranking quite well, and with a main key word you can get up around the 7th or 8th page pretty soon too, but improvement for that then slows rather dramatically. It really is a welcome relief; not being able to crack 1,000 really was a frustrating experience. I know there were all sorts of pundits dismissing this phenomenon, but for us it was quite real.
Well, I was really asking the opposite thing, but the mod compounded the threads. So nobody is seeing sandbox effects new ages of long established sites?
>>So nobody is seeing sandbox effects new pages of long established sites?
I haven't noticed it. Added two pages to a 6-year-old site Sunday; they're ranking about where they should be today.
Whoops - sorry for crossing up the topics here. I haven't personally seen this sandboxing of new pages on established sites, but several people here have reported it recently.
I've been wondering about what might trigger the delayed ranking effect for some new urls but not all? Might it be the URL (a new directory?) Might it be some other factor on the new page - such as not using the same page template? The presence of new outbound links that Google's algo needs to examine for trust purposes? Something as simple as how competitive the search terms are? Could it be that you need lots of competition around you on the SERP just to notice the effect.
For established sites it seems to be the anchor text/title combo.
The usual pattern of analysis is...
Is it relevant to the site?
Is the site relevant to the theme?
Is the site trusted for the theme?
It's so predictable it hurts.
example: Photo blog. Every post entry is the name of a town or village. Not much of a link profile. One of the regions is the same word as the name of the International airport near that area ( -airport of course ). ALL URLs indexed within two to three days, except this one. It took a month to index, and still ranks at about 600+
For brand new domains, it's the thresholds set for the theme. At least if it's not that competitive, you can get in there fast. A few trusted links are enough even if others are bidding for the phrases in AdWords.
But if it's very competitive,... well, you're going to see an invisible wall at the usual positions. 180, 160, 140, 120, 100, 80, 60, 40, 20. As links age, you can get into a new league, but... it's better if you don't wait, and get more quality links instead.
The "sandbox" is still tied to "trust".
With a twist of "trust" now teaming up with "relevancy".
There's a huge offset for the thresholds of competitive / not competitive phrases.
|I've been wondering about what might trigger the delayed ranking effect for some new urls but not all? Might it be the URL (a new directory?) |
Not in our case, site is small enough that we fell into a flat directory structure.
|Might it be some other factor on the new page - such as not using the same page template? |
Using the same template and site logo, but the internal navigation is distinct to the new collection of pages in this case.
|The presence of new outbound links that Google's algo needs to examine for trust purposes? |
No outbound links at all, guess we're hogs:-)
|Something as simple as how competitive the search terms are? |
Not in this case, sufficiently esoteric and sufficiently voluminus that we would see at least one Google referal after a week or two indexed.
|Could it be that you need lots of competition around you on the SERP just to notice the effect. |
Not sure what you meant by this.
May have seen some movement today, but I'm too compulsive to check stats in the evening, strictly a morning job.
|For established sites it seems to be the anchor text/title combo. |
|Is it relevant to the site? |
The new pages are relevant to a chunk of the site, as are the titles. There aren't any links outside of the navigation, so I'm not sure how anchor text applies here.
|Is the site relevant to the theme? |
The site itself is not themed (themed a word?), ie, does not have a single theme.
|Is the site trusted for the theme? |
But you do raise a good point that's been nibbling at the back of my mind. It could be that Google thinks that the new collection of pages, a substantial expansion in depth and theme over a couple older pages, looks like duplicate content - theme wise. We've never really paid any heed to theme since we don't have one, but maybe Google is uping the ante there, and we would have benefited from at least using directories. Too much deep linking at this point to change.
|The "sandbox" is still tied to "trust". |
It may be but how did my websites on new domains gain the required trust so quickly? Wouldn't it have been impossible for them to do so?
I am seeing exactly what the OP describes on two of my sites.
One six year old .CO.UK site, roughly 1,000 pages. Prior to five months ago, new pages were indexed immediately and had a normal PR of 3. Over the last five months or so new pages are not indexed as normal. They only appear in the SERPS for very specific searches. Older pages are fine in the SERPS and have good PR.
A glimmer of hope appeared a couple of weeks ago when a batch of the newer pages appeared normally in the SERPS. They are still doing fine. But some of the other new pages are just not there.
I have absolutely no idea why. Given the "glimmer of hope" above I am now leaving the pages alone and hoping they will appear in the SERPS soon.
|It may be but how did my websites on new domains gain the required trust so quickly? Wouldn't it have been impossible for them to do so? |
The reason I mentioned relevancy, trust and themes is because different themes -- and different keyphrases within those themes -- trigger different thresholds for the SERPs.
For some it's enough to have your site crawled.
For some a single quality link is enough for a top 20 position ( see below ).
For some you still need the propagation of at least 20-30 regional/thematic/trusted hubs.
The less competitive, the lower the threshold is.
And by competitive I mean advertiser competition, not search traffic or number of pages in the index. With consideration of the competition involved because of overlap of interests for the individual words that build up the key phrases.
An over simplification of the issue, consider Google has segmented the SERPs into different stages of trust / importance, and only within these stages would relevancy decide your position. Meaning the top 1-20, 21-40, 41-60 ... 80, 100, 120, 140, 160, 180 are separate "leagues" in generic web search. ( not counting special, injected channels, filters, demotions ). If it's highly competitive, the count goes down all the way to the last position of a two word phrase, where it's cut off at let's say, 200-300 results, as if there weren't more.
With more trust, you can enter this race or jump up some levels. ( OK, we call it trust, but it's in fact 'local trust' ) You could get 10,000 blog post links with no effect on your ranking. And a single quality link that'd lift you to the top 20. From then on it's about relevancy and volume.
So... yeah, for less competitive phrases, there may be only two leagues ( top 20 and the rest ), while for more competitive areas, there are several. In such a mild environment you can get a brand new domain ranked within a week, I did so too this month.
Both trust and relevance are in fact numeric parameters, can be boosted with quality links. Although relevancy has so many cross references between analyzed words and phrases, you can only count on your intuition.
For all three ( trust, themes, semantic relevancy ) it's the inbound links, internal links ( for themes/relevancy add titles and content analysis ) that decide where you are compared to others.
If a specific theme, or entire sector is out of Google's understanding/monitoring/Keyword auctions... it's easier.
They have refined many things the last year, including the threshold in any of the sectors which decided what sites could appear for 2, 3, 4 word queries. If the query is short, but not generic, they let you on the SERPs much faster. This wasn't so before.
|May have seen some movement today, but I'm too compulsive to check stats in the evening, strictly a morning job. |
Nope, still getting the cold shoulder from Google on the new pages. Almost wish we'd started a new domain for them, after reading about the site with new ages that have been left behind for many months.
|The less competitive, the lower the threshold is. |
Agreed, and I think this may be extremely important. None of the sites I have launched are in competitive areas.
I haven't started any brand spanking new sites recently; adding new pages to existing sites has gone extremely well - I've had new pages up and ranking in the top ten within three days. Does seem that the older the site the faster that new pages get up and running.
|Does seem that the older the site the faster that new pages get up and running. |
Well, it always worked that way for us until recently. This morning I looked over our stats, and the new pages that have been in Google around two weeks without a single referal are getting search traffic from Yahoo! and Alta Vista, of all the places, and I didn't even know Alta Vista was still running!
I wonder if Google is waiting for a quiet period of a few days or a week before they take the new pages seriously. Due to template tweaks, additions and editing, I'm not sure two days have gone by without all the new pages getting uploaded again.
Not only "older sites" ( stable URLs ), but also, "older incoming links".
And not "older", but "more trusted".
Link age calculation is a part of trust.
There are huge differences between established sites and established sites.
I mean a domain could be hosting a site for 8 years with little or no inbounds links, and face the same "sandbox effects" for any and all new pages added. While a relatively new domain would jump ahead within a month or two for the given keywords.
Again: this is about inbounds links.
Big surprise huh?
If your site has a healthy link profile with some quality, long time in place linkage, with a variety in relevant anchor text, your site will propagate trust, relevancy and its theme to any new page ( again, given the proper anchor text in your internal nav ). This makes top ranking with a new page possible within about 4 days for regular sites. For news related / blog sites, the turnaround might be even faster.
Established sites are not about domain age, number of links, PR or whatever. 'Established' in my SEO vocabulary translates to well earned links from authorities, hubs, variety of anchor text, a recognizable theme... pages soaking in relevancy and signals of trust.
A new page on such a site is a piece of cake to introduce.
Quite a rewarding effort.
|'Established' in my SEO vocabulary translates to well earned links from authorities |
Yes, but we meet all of that and then some with sugar on top:-) I understand where you are coming from, but I'm pointing out that for some apparently *new* reason, introducing new pages is no longer "cake" for us, or at least, not the last batch.
The new pages aren't suffering from low visibility in Google, they are suffering from a penalty. I just don't know if it's a "new" penalty, a "too many pages in the root" penalty or a "stop fooling around with the new pages" penalty. Or, it could be something I haven't thought of, which is why I started this thread.
Oops, an update. Some of the new pages did start picking up Google traffic yesterday, when I checked this morning, I just looked at the ones I thought would be the most promising because of their higher traffic levels DUE TO INTERNAL LINKING. Apparently, Google may be bothered by their sudden appearance in the site template. The new pages that weren't given prominent links are ranking top 10 for some phrases.
I'm noticing for our 10 year old site that google is not only looking at the anchor text in an IBL but also what is around it. We're ranking on a few terms that aren't anywhere on a particluar page, but a page that links to it has that key phrase on the page but NOT in the link itself.
Bewenched, I have also seen the title element of the linking page have an effect. As Miamacs summarized, it seems to be trust, themes and semantic relevance. Google's semantic relevance algorithms seem to be maturing to take in a good bit of the "surround".