| 1:56 pm on Jun 20, 2002 (gmt 0)|
>>>>Anyone see a problem with that?
Yes...I wouldn't do it. Imo.
| 2:12 pm on Jun 20, 2002 (gmt 0)|
With just two domains I don't think it's automatically a penalty, but along with agerhart I probably would avoid that now.
| 2:15 pm on Jun 20, 2002 (gmt 0)|
Okay. Could I be bold and ask what you would do?
I have 2 sites both owned by the same company and on a similar topic.
SiteA could just link to SiteB with no return link, is that better or am I overdoing the caution you're instilling in me ;)
| 2:23 pm on Jun 20, 2002 (gmt 0)|
we have 4 sites on widgets. Plus 1 company page. They are all concerning different, but related, aspects of our industry.
Are you guys really saying that we could lose ranking on google simply because we link between these sites.
Surely a cheap regional widgets site could link to a hairy widgets site without penalty? Even if they are both created by the same company.
| 2:27 pm on Jun 20, 2002 (gmt 0)|
> Could I be bold and ask what you would do?
If these were the same two sites as in the other thread [webmasterworld.com] then I personally would link only from the category page of the general site to ProductA.com, and from every page of ProductA.com back to the general site.
As far as I can tell from Google's listings up to now, a hub can have lots of spokes, but the spokes shouldn't all link up.
The problem, Nick, is that you don't know that the relatively careful hub and spoke approach that has served me well up to now won't land me a PR0 at the next update, or the one after.
shanz, I've seen enough penalised sites to be very wary of linking them all to each other, whether they're created by the same comapany (or on the same IP or class C) or not.
| 2:39 pm on Jun 20, 2002 (gmt 0)|
I'm dealing with a very similar situation now, with the site I asked about with two domain names pointing to the same site. Now they're separate, on different IPs, but every page on brandname.com links to mainsite.com - and some link back. With one look I know it's headed for trouble. I've seen some sites interlinked and get by with it, but it seems very risky.
| 2:40 pm on Jun 20, 2002 (gmt 0)|
|I personally would link only from the category page of the general site to ProductA.com, and from every page of ProductA.com back to the general site. |
Thanks callum, that's exactly what I'll do. Nice to have a clearer view on this. I've not worked on a project like this before.
| 2:51 pm on Jun 20, 2002 (gmt 0)|
OK, but please promise me that you take the warning seriously.
Two days from now I could be wiped from Google, and I don't want anyone thinking that I'm suggesting "this will keep you safe in Google"; I just can't say that.
| 2:51 pm on Jun 20, 2002 (gmt 0)|
> I personally would link only from the category page of the general site to ProductA.com, and from every page of ProductA.com back to the general site
You would? Linking from every page of one site to another site? Wouldn't that stick out?
You are the Google expert here... but darn, that's what I thought would be risky.
BTW: where's the logic in linking from every page?
| 2:53 pm on Jun 20, 2002 (gmt 0)|
Logic, none really other than a 'visite our parent site' type thing. I guess that could be ammended to only include the index page and about us page.
Might be safer still that way...
| 3:16 pm on Jun 20, 2002 (gmt 0)|
to revert to one of your areas of expertise - multi-lingual sites and the interlinking subject of this thread:
Do you know of any sites with basically the same content and layout, but all in the local language of the tld, all linking from one subject page in one language, to the other language subject pages. That is a strict one-to-one page linking from one language site to the other.
That is, from blue-cars page on site.com
to blaue-PKW on the page of site.de etc,
and that for every local language subject page and vice-versa,
without incurring Google penalties?
I am considering this as a service to the visitor, who would look for something in English first, because generally there is more information available in English, find it on my .com site, and to his surprise he finds there is also a version available in his native language.
| 3:51 pm on Jun 20, 2002 (gmt 0)|
Vitaplease - all kinds of sites do all kinds of things which have brought some (or rather: many) other sites in trouble with Google.
No hard rules out there...
Only some months ago I would happily recommend setting up multilingual content the way you described. I don't deem it neccessary to interlink every page though.
Still: the duplicate issue should be neglectible in such a scenario. The crosslinking issue however is a serious one.
My current strategy is therefore, as mentioned in the other thread on Nick's two sites, to interlink the sites , but also to develop the sites independantly.
Each site of the network should get as many incoming local links, which only go to that specific site of the network as possible.
Exactly that is one of the major advantages for SE promotion purposes in such a setup anyway, and not only a way to avoid crosslinking penalties.
I am very wary nowadays anyhow with heavy interlinking. I somehow still visualize the whole problem as maps of the web, where isolated closed rings of sites are easily recognizable.
No site is an island - no pair or group of sites should be either.
| 3:59 pm on Jun 20, 2002 (gmt 0)|
My perception is this discussin has been about the "risks" of having every page on site A give a link to Site B. Aside from the added risks of the multi-page approach, what's the benefit (to site B) above and beyond the approach of having only the root page of Site A linking to B?
I would have thought that the PR contained by Site A's root page is mostly redundant to the PR that the second layer pages of A site. But this discussion seems to infer otherwise?
| 4:03 pm on Jun 20, 2002 (gmt 0)|
T-Rex, as Nick said in msg 10 it's not so much a PR thing for him, but about branding and usability.
| 4:16 pm on Jun 20, 2002 (gmt 0)|
I confused. Suppose you have your web design site, Site A. Then you have Site B, which you created. Each site has separate incoming links that are relevant.
The issue:Suppose Site B has a copyright footer on every page that links back to web design firm, Site A?
Sounds legit but should I not risk it because Googlebot might not be smart enough to figure it out?
| 4:21 pm on Jun 20, 2002 (gmt 0)|
I want siteA to have as much PR as possible, so linking from the index and about page of siteB to siteA's index should belp but not hinder.
SiteA will only link to siteB form /catagory/index.html
although it seems a shame not to be able to infer more PR to siteA
| 4:28 pm on Jun 20, 2002 (gmt 0)|
The thing to clearly avoid is every page of every site linking to all the other sites.
Let's say I had these six sites, (For Branding Reasons like you said)
I would not want a link from on every single page of ProductQ.COM to go to all the other domains. Clearly this could be for marketing, but what it is doing is creating a PR0 nightmare waiting to happen. Be smart about how you interlink. Allow CompanyABC.com to be your hub if you will, if you still want to do your interlinking with no penalty possibilities then do a redirect link and tell Google that the redirect page should not be crawled (in robots.txt). This means you will not pass PR and Google will not consider you for PR0. This will allow humans to follow your link nest, but Google will not see the interlinking.
The rule at Google is; Do not pass PR, do not go to Jail. Collect $200.
| 4:31 pm on Jun 20, 2002 (gmt 0)|
> > I personally would link only from the category page of the general site to ProductA.com, and from every page of ProductA.com back to the general site
> ...that's what I thought would be risky.
Yes heini, I should probably explain what I mean there. As I dig myself deeper and deeper into the hole of looking like I'm telling people they should do this, I want to be clear that this is very much my personal take; I do not advise people to do this, nor do I advise people not to. We each have to set our own safety margin.
'Cross linking' is left purposely vague by Google. First the easy part...Even with less than a half dozen domains, all linking to each other, I've seen people get PR0 and I'm 99% sure of cause and effect. Every hub and spoke relationship Iíve seen has no PR problem from it, even with plenty more than 50 domains.
In the quote above, I'm only describing hub and spoke between two sites. It's a common feature of link relationships and it's not a common feature of PR0 sites, or at least those I've looked into.
The hard question, for me, is where the domains come into this. It is safe to have heavy crosslinking within one domain (i.e. fully qualified domain name, not counting a.domain.com and b.domain.com as the same), as the ubiquitous navbar demonstrates. But not several domains. This gets away from the page based nature of Google, but that was always a possibility (they cluster results by domain anyway).
As far as I can tell, heavy cross linking between just two fully qualified domains isn't a problem, but I get nervous about that and wouldn't want to go as far as to write "Marcia, that sounds OK as far as Google's been up until now", even though I'm thinking it.
I've often read advice about getting links from sites outside the collection. Even if it's not "the PR0 cure" (I don't think it is) it's sensible for several reasons.
martinbuster, I'm wary about saying so, but I don't see a real problem with that so far, if each site links back to just one. I would be very afraid to have more than two domains for different languages linking all to each other. That's a pity, but we need to be aware of these things.
> I somehow still visualize the whole problem as maps of the web, where isolated closed rings of sites are easily recognizable.
Nicely said. For 'sites', it seems that we can read 'domains'; so far at least.
As heini and Nick said, it's not about getting more PR. Heavy cross linking is, in my opinion, a poor use of PageRank.
|brotherhood of LAN|
| 4:45 pm on Jun 20, 2002 (gmt 0)|
Thanks for the clarification on your approach ciml. I definetely think that its a less risky way than the initial way that Nick suggested.
Googles concerns will also be slightly different, i.e. if every site cross linked, then it would travel the web the long way, through the hoops we tell it to go through.
We can only provide so many hoops...and I guess the hub and spoke approach as ciml puts it is a way to link I would feel OK about.
| 4:48 pm on Jun 20, 2002 (gmt 0)|
> The thing to clearly avoid is every page of every site linking to all the other sites.
What about having several sites, in different language? (each site is just a translation of the main English site).
It would be fine that you can just click on a flag to switch to see the page in that language.
Does it mean that you would be penalized?
| 4:51 pm on Jun 20, 2002 (gmt 0)|
Angiolo: check out msg #12
| 5:00 pm on Jun 20, 2002 (gmt 0)|
I just want to add that few weeks ago I sent an email to Google about that.
I received an automated reply, but I haven't yet received the "personalized" reply.
| 5:01 pm on Jun 20, 2002 (gmt 0)|
| 5:14 pm on Jun 20, 2002 (gmt 0)|
>I do not advise people to do this, nor do I advise people not to. We each have to set our own safety margin.
Extremely important point. There's much to lose for many folks, especially when it comes to the small company/guy/girl.
If you're building to last - be cautious. Advice as in linking for humans, not bots, is good if you need to stay on the safer side.
But then - we're here to learn advanced stuff, right? And we're all grown ups. Most of us, at least ;)
|brotherhood of LAN|
| 5:48 pm on Jun 20, 2002 (gmt 0)|
do we have any indications that it's (cross linking) a human imposed ban or just an SQL-automatic-type algo penalty?
| 6:11 pm on Jun 20, 2002 (gmt 0)|
We're talking about the automatic PR0 penalty here. As far as I know, it's still the case that human penalties cause the "removed from index" behaviour (grey PR, no spidering) instead of PR0.
Recently, though, I get the impression that Google engineers sometimes remove the automatic PR0 penalty by hand. Maybe they now give them by hand, too?
One more disclaimer: We're talking about penalties here because some kinds of link relationships have been found to cause PR0. Penalties are rare; most people with a grey ToolBar PR graph have some kind of spidering problem, bad robots.txt or just haven't waited long enough to be included. The vast majority of Webmasters should sleep well at night.
| 6:27 pm on Jun 20, 2002 (gmt 0)|
I'm no expert, but would it be a lot of trouble to locate all the sites under the same domain, and then point the secondary site(s) to the proper URLs? For example,
www.widgetsinc.com <- your corp site
www.widgetinfoforyou.com <- a nice site for info about widgets, "sponsored by widgetsinc.com"
www.widgetsinc.com/widgetinfo/ <- info site moved under your other domain.
www.widgetinfoforyou.com <- forwards to above URL.
Seems this would make your "cross-linking" no more than navigation linking within your site, so no penalty, right?
Once again, not an expert, so I could be wrong.
| 6:40 pm on Jun 20, 2002 (gmt 0)|
I don't think you're wrong Dan.
That was my first idea for this project. But, I realized that I could significantly increase the clients chances of being 'clicked' with two urls stuffed with content instead of one.
| 6:45 pm on Jun 20, 2002 (gmt 0)|
<<<<<<<<As heini and Nick said, it's not about getting more PR. Heavy cross linking is, in my opinion, a poor use of PageRank. >>>>>>>>>>
ciml, I'm still confused. As in Nick's post 17, it is about PR. In post 13 it was pointed out the concept that PR from the index page is probably mostly redundant to PR from the second level pages. (unless you have a specially, mature site with lots of outside-targeted additional keywords on the sub-pages) But I was under the impression it is a new site and that Nick wants to maximize PR transfer to the develper site. So the question in post 13 still stands.
| This 84 message thread spans 3 pages: 84 (  2 3 ) > > |