|Can you shed some lights as to why having a subdomain is less beneficial from an SEO stand point than going the directory route. |
Because , quite simply , most "non geeky" or "non pro" webmasters or site owners will unthinkingly link to your blog using your first example ..ie ;http://www.example.com/blog ..and hardly ever ( unless they appreciate what links are ..or you ask them specifically ) link to your second example blog.example.com ..
that is to say they will say .. I read a real good blog at this site http;//www.example.com/blog ..and hardly ever ..see what "X" is saying at http//blog.example.com" ..you would be amazed at the number of people on the web ..with sites who dont understand what a sub domain is ..
IMO ..until the average webmaster IQ goes up 20 points ..thats the way it will stay :(
[edited by: Leosghost at 9:54 pm (utc) on June 5, 2009]
SEs may see the subdomain as a separate site so even though you cross-link, the main site may not accumulate as much PR as the it would with the subdirectory route.
|will unthinkingly link to your blog using your first example |
So set up a redirect.
If there's a profitable estore already on the domain, protecting that should be the highest value here.
If there are technical reasons for preferring blog.example.com instead of example.com/blog, those should be given serious weight. The technical issues are real and permanent, whereas the SEO differences are uncertain, and disadvantages (if there are any) can be overcome by savvy link development.
The technical reason why IT prefers blog.example.com is because the blog will be hosted on a PHP box, while our site runs on Java servers, so it is is easier to just have different DNS.
But technically, it is possible to setup www.mysite.com/blog to pull data from a server hosted under a different DNS. This is called reverse proxy. The only downside is that the data must take two trips instead of one before reaching the client.
I did some research, and here is my analysis of the situation. Let me know if you have other opinions:
After further researching the topic of subdomain vs. directory for the blog, I strongly feel like we should implement the reverse proxy solution. Here’s why.
The main reason why we want to have a corporate blog is for Search Engine Optimization purpose. Indeed, there are 3 main SEO benefits of having a blog:
1) A blog is typically more regularly updated and has article covering more topics than our eCommerce site. As a result, the content from the blog allows us to appear in the search engine results for queries for which the main site’s content wouldn’t typically rank, thereby increasing the reach and exposure of our site in the search engine.
2) The fact that the blog has regularly some new articles added to it weighs in significantly in the ranking algorithm used by search engines. “freshness” of content contributes to good organic rankings, and positively affects the domain of the site on which it is published.
3) Good blog posts are likely to attract links from other sites. Inbound links carry “link juice” – meaning that the site linking to us is endorsing our content and designating it as a valuable read for its audience. This is plays a tremendous parts in the ranking algorithm of Google and other search engine (in fact, the Pagerank concept calculated based on inbound links was invented by Google). The more inbound link we get to our content, the more authoritative our site will appear in the eyes of the search engines, and the more likely we will be to appear at the top of the search engine result page.
The problem if we implement the blog under a sub domain is that none of these benefits will materialize, because search engine typically treat subdomains as a totally different site than the root domain.
As a result:
- oursite.com – which already enjoys a pretty high pagerank – will not carry its pagerank to blog.oursite.com. It would be like starting a new site from scratch, with a zero pagerank, and we would have to spend a lot of efforts getting it indexed in google. With no initial pagerank, it could take a good year before our blog would start ranking for any search query, because it would be seen as a new site with little recognized authority.
- With blog.oursite.com being seen as a different site than oursite.com, the “freshness” factor brought by the blog will not positively impact oursite.com
- Lastly, inbound links from other sites to blog.oursite.com will not contribute to raise the Page Rank of www.oursite.com
In conclusion, I strongly recommend that we take the extra step of implementing the reverse proxy if it is technically feasible and not too costly.
[edited by: eelixduppy at 2:41 pm (utc) on June 10, 2009]
[edit reason] removed specifics [/edit]
|if we implement the blog under a sub domain is that none of these benefits will materialize |
You seem to be assuming that there will be no cross-linking between the blog and the main site.
If you cross-link normally, the PR and every other benefit will flow just fine. The worst you can expect is that it might take a bit longer for the blog on a separate subdomain to "find its feet" SEO-wise, but even that is a "maybe" and it can be overcome with savvy link development.
As for the reverse proxy stuff, why make life more complicated than it needs to be? That's just a recipe for trouble. To me it would make a lot more sense to invest the energy into link development.
Thanks for your response buckworks, but I'm not quite convinced.
First, we will probably not link from our ecommerce site to our blog. We dont really have much room on our ecommerce site for editorial content, and I dont see where or when we will ever link to our blog. Therefore, our main site will not pass any link juice to the blog.
We will, however, link heavily from the blog to our main site - but what good will it do, if the blog has zero pagerank to begin with.
Second, we dont want to engage in any link building efforts for our blog residing in a subdomain. We're spending enough time as it is trying to optimize our main site (which exists under 3 different TLD & languages for each market we service), we don't want to divert resources in order to develop another domain.
We're hoping that the blog will garner inbound links by itself, just because we will post good content. We'd rather have those occasional inbound links work towards raising the profile or our main domain.
Therefore, it sounds like the SEO benefits for our main site will materialize immediately if the blog is in a sub directory, whereas if the blog is in a sub domain, we will have to first wait months for the sub domain to earn some degree of legitimacy and authority in the eyes of the search engine before it will be able to pass any SEO benefits to our main site.
Technology should serve business goals, that's why I want to support the reverse proxy solution, if it is in fact reasonably simple to implement (which it is from my understanding. Just an Apache configuration setting).
Buckworks is correct that the blog will receive and pass PR if you cross-link effectively. There may not be a 1:1 relationship in link value however. In the near past the SEs (G in particular) treated them as separate entities but those lines are gray now. In a subdirectory every link (with value) is counted towards the domain.
Also, WP is an excellent blog but in your case, why not look for a Java-based blog? It may have less features but it will be a lot simpler to implement and maintain.
Technology is not an issue here. If we insist, IT will do the reverse proxy. It is fairly simple to implement. This is really besides the point. I just want to make sure that my argument are valid so as not to fight a battle not worth fighting for.
But from what I've read, I still think we will be better off setting up the blog as a subdirectory...
I should qualify two statements:
"In a subdirectory every link (with value) is counted towards the domain."
Provided the internal linking structure is optimal.
"There may not be a 1:1 relationship in link value however."
Meaning that if every page with PR is not linking to the main site then that page would not pass PR.
If you plan it correctly though, buckworks is on the money.
Thanks for your response Excellira.
I realize that the blog on a subdomain will pass some of that subdomain's Page Rank to the main site. The problem is that the subdomain will start with a zero pagerank.
And like I mentioned, we don't have the possibility of raising the PR of the subdomain by linking to it from the root domain, because we don't have much room on the root domain for editorial content where we could link to the blog.
Therefore, before the blog on a subdomain can pass any pagerank to our main site, we will have to wait first for the blog to build some pagerank of its own, which could take a while and some resources.
I would much rather have the inbound links the blog will naturally generate contribute directly to the pagerank of our main domain.
I assume I will get a much greater SEO boost for our eCommerce site with links pointing to www.oursite.com/blog/name-of-post instead of blog.oursite.com/name-of-post (even if the artcile links back to oursite.com)?
|And like I mentioned, we don't have the possibility of raising the PR of the subdomain by linking to it from the root domain, because we don't have much room on the root domain for editorial content where we could link to the blog. |
Regardless of where you host it, if the blog is important to your business why wouldn't you make space for a simple link to your blog? How much space does a text link saying "companyname blog" really take up? Do you have a footer section on your site?
[edited by: MadeWillis at 4:10 pm (utc) on June 10, 2009]
I typically prefer the subdirectory route myself but I like to expose as many of the variables as possible so that you can make an informed decision. We haven't delved into branding, trust, and other factors but this is a deep topic that has been discussed elsewhere on WW.
PR works on a page-level. So, unless the page is linking to the home page or another internal page, it would not pass PR to the other pages on the site. So, if your blog was linked to from your ecommerce site's navigation then it would receive PR fairly quickly regardless of its location.
Also, if you're planning on getting posts into Google news you'll need a three digit number in the URL (see: [google.com...] ). This might work:
Of course, we will link to the blog from a few places. Definitely.
But the point I'm not getting is why would we want to worry about trying to raise the PR of a subdomain, when we could bring it right away to the already high PR level that our root domain currently has, by setting up our blog as a subdirectory when we launch it?
If we launch it as subdomain, even if we give one link to it from our root domain, it will take months before the content of the blog starts ranking for search queries and gain visibility. One good link from our root domain will not be enough to give it enough of a boost, don't you think?
The PR flows to wherever you link. If you don't link to your blog it won't automatically receive PR from an internal source such as your home page. So, regardless of its location, if you link to it in your navigation or elsewhere across the site you'll pass PR to it.
|The PR flows to wherever you link. If you don't link to your blog it won't automatically receive PR from an internal source such as your home page. |
Exactly. Your blog won't "start" with any PR just because you host it in a subdirectory. You must link to it one way or another. The crawlers need a way to find the page. Even one link may not pass the PR you expect. A company blog many times deserves to incorporated into the navigation stream somehow (header nav, footer nav, etc)
They way I look at it is will the content be simialr (on the same topic) as the content on your root domain? If so, then a subdirectory is probably the best choice. If the content will stray from that on your root domain then create a subdomain.
Take a look at major sites, like Google, as to how they use subdomains.
I see your point. I was maybe too quick to assume that our main site's PR will generously flow to the blog's subdirectory without much effort.
But further to MadeWillis comment, I think the subdirectory will be better suited, because the content of the blog will be quite closely related to that of our site. Our eCommerce site is about Travel in Europe, and so will be our blog. So I assume it makes sense to have them closely tied.
|The problem is that the subdomain will start with a zero pagerank |
So would example.com/blog - unless it has some links.
|we will probably not link from our ecommerce site to our blog. ... I dont see where or when we will ever link to our blog. Therefore, our main site will not pass any link juice to the blog |
Then either blog.example.com or example.com/blog would be at a disadvantage.
You seem to be assuming that the blog would somehow acquire instant PR merely from being located at example.com/blog. That assumption needs a closer look.
The key issue here is linking, not location. No matter where you locate the blog, it will not have more PR than it deserves based on its links.
|if your blog was linked to from your ecommerce site's navigation then it would receive PR fairly quickly regardless of its location |
Buckworks - thanks for pointing out the flaw in my reasoning.
Now, if we put a link to our blog in the footer of our main site for instance, it will pass pagerank from every page of our main site to the blog in the same fashion, regardless whether it is in a sub directory or a sub domain. Correct?
Therefore, the question of hosting it as a subdirectory or as a subdomain is not as crucial as I tended to think initially... ?
Yes, though site-wide links may not be valued as highly but you also have the opportunity to pass blog PR to your ecommerce pages by incorporating links into the body content. If you blog about travel to Venice you could place links in the body content (or at the bottom of the post) to the Venice travel page. If in the bottom, you could suggest that the visitor check out your Venice travel packages. Be careful not to overdo this or you'll undermine the trust you are attempting to build with the blog.
Just to add to Excellira comments
I have site wide links to my category pages in my navigation as they are all major parts of my site. I am not being penalized for this. You may or may not wish to do this with your blog, however if it is an important piece, I would probably go ahead and do it.
If not, I would just add a link towards the top of your homepage in an effort to pass as much PR through that one link as possible. Based on your other comments, this may be the best option for you.
Carefully consider how you setup you blog and categorize your posts before you set it up. Keep the blog posts in tight categories and link those category posts back to matching categories on the ecommerce side.
Thank you very much. Those are very helpful insights.
"I have site wide links to my category pages in my navigation as they are all major parts of my site. I am not being penalized for this. You may or may not wish to do this with your blog, however if it is an important piece, I would probably go ahead and do it. "
Agreed. My comment should have stated that you should do it regardless of how much value those links pass.
|opportunity to pass blog PR to your ecommerce pages by incorporating links into the body content |
PR would flow, and the ability to create strategically worded links in context would also be useful.
|the question of hosting it as a subdirectory or as a subdomain is not as crucial as I tended to think initially |
I would agree with that.
If there is an SEO difference here it's not easy to tell, which means any differences would be small. I don't think you're at risk of a huge SEO blunder either way, and I bet it would make the techies happier to put the blog on a subdomain. Any SEO differences would smooth out as the blog gained link popularity of its own.
The biggest benefits of a subdomain used to be that you could have 4 listings on a single SERP. 2 from the root domain and 2 from the subdomian. Google has since fixed this flaw and only in very rare cases do you see 4 listings like this.
Leaving aside the Pagerank discussion, where it seems that subdirectorty vs. subdomain is pretty much neutral, what about:
- "content freshness" brought by the blog positively reinforcing the root site if the blog reside in a subdomain
- overall reinforcement of our root domain authority within its vertical (european travel)
Am I right in still believing the subdirectory solution is still our best approach SEO wise?
Oh, I just saw your response Buckworks.
I think you're probably right after all, and like you mentioned, it will definitely speed up our time to market and make us better friends with our IT department :)
Let's use a blog that will get 50 external links from different sites.
example.com [domain] benefits from the link diversity of 50 sites.
blog.example.com [domain] benefits from the link diversity of 50 sites.
example.com [domain] benefits from the link diversity of 1 site (blog.example.com).
So it depends on how strongly you believe in backlink diversity.
[edited by: BradleyT at 6:18 pm (utc) on June 11, 2009]
Really straightforward this one.
If you write good content, it will get visbility.
Aside from technical descriptions I've stopped using subdomains, although there was a time when they were very effective.
Domain.com/blog should be fine. It's much easier to remember too, because if someone remembers your domain they just carry round Blog in their head (how many people from the UK actually need to remember the domain name when the National TV station gives out a weblink, you just remember one word). Similar thing with this. Put a weird bit of syntax that might sound easy to you and me, but to the novice means they have to insert a lot of extra dots before they get the successful format that shows the blog. Even worse the average users thinks that all websites start with www. in which case you'll never get any benefit from a subdomain.
Personally I wouldn't do it just for that.
For the record, domain link juice goes back a forward whether subbed or not.
well, that was an interesting discussion, and obviously there is no obvious best solution from an SEO standpoint...
Thank you all for your input.
| This 40 message thread spans 2 pages: 40 (  2 ) > > |