Forum Moderators: Robert Charlton & goodroi
We have over 80k of pages with around 4 internal links to a page and 4 IBL's to a page, supplemented with 3rd party IBL's into deep pages. The content on each site is in in a different language and the pages are unique. Theme's have been maintained.
The extent of the changes [ about ] 3 in the last 6 months is something like this:
Very Nice Red Widget changed to
Nice Red Widget
Our restored site is indexed, but has not ranked for around 7 months.
Will frequent changes threaten to remove our site from Google's SERP's, or prolong our existing "sandbox" filtering?
I like to consider it like this: Either you have "dynamic" pages, where more or less all of the content changes on a regular basis (single-page blog or forum posts where all comments appear on that one page and edits of previous posts are allowed/common, automated constant-shift services like PriceGrabber, etc.)...
Or, conversely, you have "static" pages, wherein the only thing liable to change are the adverts and comments (and the comments can't be edited, or aren't edited often). Pages where content is placed, and where that content has a low rate of flux.
One or the other, mi amigo. Start mixing apparently static content with small "shifts" (regular changes to JUST the anchor text, alt attributes, meta tags, etc.) and you'll find yourself in a whole world of SERPs-related Hurt.
Why, you ask?
The number one thing Engines HATE are the methods deemed common to Search Engine Spammers. In other words: if you appear to the Crawler like you're trying to game the system, you're doing a horrible job of it.
It did not.
The pages ranked according to their new anchor text, as if it was the anchor text from the very beginning.
...
On the other hand, in the navigation of a site where the bulk of the pages were supplemental it took much longer for Google to crawl and take notice of the new anchors, and thus longer for the target pages to rank for it.
...
...but...
Your problem might be comepletely elsewhere.
I'd say it's in sites of different language interlinking. ( If I got that right. ) Sites of different language are almost as if they were a different theme. There's no cross-language calculation for relevance from what I can tell -- no backend dictionary -- , meaning those links are practically off-target, and pass as little "link juice" as a small vote on the exact phrase.
Pages may be same theme ( as in identified as same category, which is Travel or electronics I assume, but perhaps not ), but unless the keywords are the exact same in all languages, they are irrelevant.
TrustRank is now thematic in its effect, so yes, mentioning the sandbox ( lack of trust ) is a good call. I'm not sure but I'd say you can't create a site network in different languages anymore. If you step up your campaign you may even end up with a penalty.
- I'd get rid of interlinking between pages that feature the words nowhere but the anchor text of an outgoing link ( same meaning but different language doesn't count, Google will not translate to understand it ).
- Or choose phrases that can be featured throughout all pages in all languages ( like "mycity hotel" would be a good idea - if it weren't overspammed already -, but "mycity photos" vs "fotos" vs "fotograaf" vs. "photographie" vs. "pictures" vs. "bildern" vs. "images" ... is not. At least not for voting each other relevant. They don't )
For example, if in month # 1, you changed all of the above 1-3 times throughout the month, how long do you need to wait to see the real results of your changes? 3 months? 6+ months?
Assuming you do not have an authority website, does anyone have an estimate on how long it takes to recover from frequent changes to anchor text, title, or meta tags?
If you have a decent amount of traffic to your site, and Google has some level of respect for the site (I'd say a PageRank of 2 or 3 would be a good baseline), you're likely to get updated at least every two weeks or so. Convincing people with a variety of domains of different ages and content types to contextually deeplink to specific subpages of your site doesn't hurt if you're not a fan of the Supps.
But if you're doing all of that on the reg, then you probably have more than a 2 or 3 in PR by now, and shouldn't have to worry too much about garnering "long-tail" niche terms and pulling them from the depths of the supple, supple Supps. Not sure if you've noticed, but you can be suppified for one term and at the same time, for the same page, experience a complete desuppification for another.
Oh and, also this is a very new thing.
It may hold some surprises for anyone, no matter how successful up until now.
Well thought up community sites sustain themselves, I wish I had what it takes to build the hype ( only succeeded once in that area ), if you're good in that regard, congratulations, you know how to create link bait.
But.
A network of 80.000 pages cross-referencing each other without a significant amount of help from outside ( random inbound links from OTHERS )... is probably built like that for the very reason that there isn't, or is no way to build hype around the content.
If people's links don't nail your site at the top of the serps ( with a huge variety of anchor text that wasn't thought up by you ), you'll have to do it yourself.
And no, you really can't expect to vote yourself to the top by linking a german innermost page to an english innermost page, without them carrying relevancy for the phrases used. Not only the theme, but the phrases, the language. Especially if these are competitive phrases. And unless I'm mistaken, we're not talking about communities, but SEO. Well... this time at least, but Whitey, please correct me if I'm wrong.
From a usability point of view, it's great. The multi-language sites I oversee have links to switch between the pages , and people like this comfort.
But SEO-wise they don't really do any good.
And if they were cross-domain links, from running up sites within the same network... ( LINK network ) they may even do harm. -950 for irrelevant linking, and a hit by LocalRank filters as well. Neither is fun enough to have, so proceed with caution.
Well, I take that back. Spammers with auto blog writing software, covering multiple countries to garner adsense dollars, do this all the time.
Ha Ha Ha... tell me I'm wrong? ;)
Are these 800,000+ pages completely 100% unique sku's. Or do they over lap, such as 35 different colors for the exact same product?
80k sku's is easy. The company I work for happens to have that and more. But the BEST web design would be to put several colors and sizes of the same product on one single page.
I can see where G might frown on your site if you have a separate page for each of the following....
Blue Widgets Size 1
Blue Widgets Size 2
Blue Widgets Size 3
Yellow Widgets Size 1
Yellow.....
Thats just bad web design.
On the other hand, lets look at this from a different perspective..... type of inventory. Lets say you're trying to rank Walmarts website. Sure, they offer absolutely everything under the sun.... which in turn makes them a 'general' site. It's no wonder you don't tend to see them in the Serps, because the overall site/company doesn't market themselves for any one particular market. i.e.... No Site Theme. And when a site has no theme, how well do you think they're going to rank vs. a site who's sole dedication and driving force is applied to that ONE type of item... or item niche. Then take those competition sites times ten, cause god knows people will build it if there's money to be made.
This also doesn't include the overall linking profile. Got a general store website? Without looking, I'm guessing that your linking profile is a mile long with everything from 'Buy Ketchup Wholesale' to 'Tire Pressure Gauges'. I can imagine G getting a bit confused.
Sooooo... back to the point again.
Either a 80k+ site is either
A) Spam
B) Poorly Designed
C) General Goods
Yes, options B and C will allow the user to find what they want, but I 'think' your site needs to develop a bit more muscle, for each individual page, if you expect it to rank in the top 50..... more muscle than the average person would have to do that has an entire site centered around that one object.
Of course... it's just a guess. I suppose someone could send a G rep down to tell me I'm wrong. hahahha
But we digress -- the topic at hand is frequent changes in anchor text. I once played around in that territory. A while back, there was a period where internal anchor text was just plain magic, and i started having quite a play on one site. And then the bottom dropped out, and it was clear that I had tripped something or other. Since this was the only area I was playing with...
It took over a year for that site to recover properly, and for some terms, it never did. So here's the lesson I think i've learned: There is no one magic button for today's Google - so if I find that I'm obsessing about one factor, then it's time to back away.
