Welcome to the Q&A Forum

"Linking root domains" is a metric we use because it correlates better with ranking than your overall link count. In other words, if you have 100 links from todd.com (even on different, non-root URLs), that counts a lot less for SEO than having links from 100 different root domains. It doesn't necessarily matter where the link is from the home-page or a deep page (home pages tend to have higher PR, but just on average). What we're really interested in is how many different/unique domains link to your site.

I tend to agree with Todd that it's better to use your root level "/" URL for the home-page, as Google can get quirky when you target a deeper page. That said, if you're using a proper 301-redirect, most of your current link-juice should be consolidated. Ideally, I'd use the root level, but your current configuration may not be causing any serious problems (and switching always carries a little short-term risk).

I checked one of your campaigns, and it does seem like the 301-redirect is working properly. I'm also not seeing any evidence of links to the "index.htm" version or other issues. I don't see evidence of both version sin Google's index. Not sure exactly what's going on here, but I'll run it by the support team. I don't think you have cause for concern.

This is a warning in our system that often confuses people, and we're actually looking to rewrite it. In many cases, a canonical tag SHOULD point to a different URL. By their very nature, many canonical tags are on the non-canonical version of the URL. We just warn people in case they made a mistake and created a mis-match they didn't intend to.

Since this is more of a support question, feel free to contact me through the Private Message system with the URL(s) in question, and I'll take a look.

Honestly, I think this kind of thing happens fairly often. The one paid link Google will never catch is the one that we agree to at the bar at the end of the day at SES/SMX/Pubcon/Mozcon/etc. On a less manipulative level than paid links, the best exchanged links come from relationship-building. If there's a site that's relevant to your site and where a link would benefit the users, reaching out to that site is just good, white-hat link-building, IMO.

I guess the problem comes in when you try to systematize it. Practically speaking, you end up with a lot of people looking for low-hanging fruit and submitting junk. Plus, as soon as you build a link network (no matter how good your intentions), you have to market it - once you market it, it's bound to get on Google's radar.

I think there are sites that skirt the edge of this, but it's tough. MyBlogGuest is a great example - it's a marketplace for guest bloggers, and often people guest blog for traffic/links. The site is just setting up that exchange and relationship, though - it's not directly driving link purchases or exchanges.

I'm not entirely clear what the nature of the non-paginated pages is, but the canonical tag is probably a decent solution here. You may actually want to leave the crawl paths to those URLs open for a bit - Google won't process the canonicals unless they crawl the URLs. See my recent post on that subject:

http://www.seomoz.org/blog/logic-meet-google-crawling-to-deindex

Pagination is a bit tougher. You've got a few options:

(1) META NOINDEX pages 2+ (tends to be pretty effective, but depends on the nature of the pages)

(2) Use Rel=prev and Rel=next. This is tough to implement, but is recommended by Google. If the pagination isn't massive-scale, it works reasonably well.

(3) If the pagination is controlled by URL parameters, indicate them in Google Webmaster Tools. I've had mixed luck with this, and your examples wouldn't work ("/P6" isn't going to come up as a traditional URL variable).

When you say that it's "not linked to from anywhere else," does  that include internal links or just inbound? If it has no internal OR inbound links, then it hardly matters either way. If it gets traffic but has no inbound links, then I'm inclined to agree with Ben - use the canonical tag. That way, the page can "live" on both sites/domains, but only one of them will have search value.

I'm actually looking to take two blogs and consolidate them into one brand new domain, and I think I may use the canonical tag for a couple of months first and then 301-redirect them. In that case, though, it's because I'll eventually shut off the other domains. If there's value to having the page exist (for users) both places, then the canonical is a solid, long-term solution.

Use the URL-encoded format (with "%20" in place of spaces) for your canonical tag, since that really is the canonical version of the URL. It's probably not a huge issue, but that should be a bit safer and more consistent.

I'm going to generally agree with (and thumb up) Mark, but a couple of additional comments:

(1) It really varies wildly. You can, with enough duplication, make your pages look thin enough to get filtered out. I don't think there's a fixed word-count or percentage, because it depends on the nature of the duplicate content, the non-duplicate content, the structure/code of the page, etc. Generally speaking, I would not add a long chunk of "Why Buy With Us" text - not only is it going to increase duplicate-content risks, but most people won't read it. Consider something short and punchy - maybe even an image or link that goes to a site with a full description. That way, most people will get the short message and people who are worried can get more details on a stand-alone page. You could even A/B test it - I suspect the long-form content may not be as powerful as you think.

(2) While duplicate content is not "penalized" in the traditional sense, the impact of it can approach penalty-like levels since the Panda updates.

(3) Definitely agreed with Mark that you have to watch both internal and external duplication. If you're a product reseller, for example, and you have a duplicate block in your own site AND you duplicate the manufacturer's product description, then you're at even more risk.

Just adding to previous comments, both a 301-redirect or canonical tag should work here. Traditionally, we tend to suggest a 301-redirect for critical pages, but the canonical can sweep up other variants of the home-page (and it's common to have variations, like tracking parameters, https: versions, etc.), so I've gradually moved that direction.

In your case, the trick is that you're driving other content off of "default.asp", such as:

http://www.etraxc.com/default.asp?demo=part

I'm not clear if that's a unique page or if it's also a partial duplicate. If those pages are unique and you use the canonical tag, you'd knock those pages out of the index, so it would probably be the wrong choice here. If those are unintentional duplicates or low-value, then canonical is probably a good bet.

Even if you fix your internal links (which is the right thing to do), it's still good to sweep up the bad copies in Google's index, so I'd implement one of the fixes.

Agreed with Valerie that step 1 is to turn off those directory listing pages - that can be a security issue and you don't necessarily want people to see/access the whole list. Also, make doubly sure you don't have any internal links to that directory (Google crawled it somehow).

Generally, Robots.txt should prevent crawling, but it's not foolproof, and it's pretty bad about removing pages once they're indexed. If you can block the page from browsing and return a 404 for the root page, that should be fine. The other option would be to have the page removed in Google Webmaster Tools. You could request removal for the entire folder, but I'm guessing that you may want the actual PDFs indexed.

MozTrust isn't quite a measure of "trust" in the sense of link quality, per se. It was designed to mimic TrustRank, a factor we believe Google uses to supplement PageRank. What MozTrust and TrustRank (theoretically - we can't see in the black box) do is to take a set of trusted seed sites and determine how close in the link-graph any given site is to them. If your site is closely linked to sites like CNN and WhiteHouse.gov, for example (I'm honestly not sure what's in the seed set these days), you're more likely, on average, to be a trustworthy site.

Of course, as with Google, this is just one of hundreds of potential ranking factors. One measure we've found useful in some cases is the ration of MozTrust to MozRank. If you have ton of raw link-juice (MR) but your trust level is very low (MR), then it's often a poor quality sign.

We tend to use DA/PA more than MozRank these days, as they're based on a more complex model. Still, none of these factors look at the actual quality of the linking site or the "spamminess", per se. MR/DA/PA are primarily measures of the strength of your link profile, and MT is just one measure of potential quality.

It appears that Penguin tweaked multiple factors in the Google algorithm and may even be use them in conjunction, so even if we saw Google's own data, we probably couldn't pin it to just one thing.

It's always tough to tell without seeing the site, but I agree with Nakul that you should be ok - we have a "notice" level message that sometimes gets overzealous about canonical tags. It's really just a heads up - not an error, per se. We're actually re-evaluating whether that's a good way to handle the situation.

The 301-redirect that Mark and Nakul discuss is probably your best bet here, but if that's causing you implementation problems, you could use the canonical tag on your home-page (in the section):

That will help sweep up any duplicates. It is best to link consistently to the root version, though (without the "index.htm"). FYI, you've got another weird duplicate in Google's index:

http://www.twinbytes.ca/index_from_mobile.htm

Agreed - the "canonical" shouldn't just be a tag - it should reflect the actual link structure of your site, or it's not going to be as effective. Adding the additional keywords to the URL could slightly boost the profile pages, but there's always some risk to making a massive change like that. See my post from last year:

http://www.seomoz.org/blog/should-i-change-my-urls-for-seo

It's not instantaneous, but yes, you should see that number drop over time.

I'd generally agree - it's usually not going to be worth the time/effort/money, AND the links could actually be helping you. One caveat, though - IF a scraper is outranking you (and it does happen), that's a much different story. The low-quality copies will just get filtered out, but I'd definitely monitor your ranking for new stories and make sure someone else isn't using your content to beat you on your own search terms. That's a much different situation.

I'm a little confused, because you gave different URLs for the target of your canonical tag and your "target page", so I just want to make sure we're referring to the same thing here. You don't typically need a canonical tag on the canonical version of the URL. Technically, you shouldn't put one there (Bing has specifically said they don't want that, but Google has eased up on it), but practically, I've rarely seen it cause any problems.

In other words, I wouldn't lose sleep over it

Just make sure that the "target page" doesn't actually represent multiple URLs. I've seen some people get confused on that. Typically, having a canonical tag on your home-page can help sweep up variants you might not think about, so I think it can make sense to have one, even on the target. In most cases, though, it's not necessary.

You have 101 unique root domains, and 95 of them are all the same spun article? Yeah, that's not good. If you can get the network to drop the article, I'd do it now. If you can't, build up some higher-quality, diverse links as fast as you can. If those numbers are accurate, you are in danger.

If you got the warning, you need to pay attention to it. 4/18 wasn't the over-optimization drop (not to say you aren't over-optimized, but that hit this week). There was a 4/19-ish Panda update (3.5), but that seems relatively minor. You more likely got caught up in some of the new link culling.

Having 40+ links per site doesn't make them the culprit. I'd be much more inclined to see what the 3rd party did and if they did something massive scale. Get a list from them, if you can (since it takes us a while to update the data, and we don't see all low-value links).

Unfortunately, we are experiencing a delay, due to an attempt to expand and improve the crawl. I don't have an exact ETA, but this is our TOP priority and we definitely regret the issues people are having.

I would double-check that Google has indexed and re-cached the page with the link on it. It's not instantaneous, in practice. Also make sure the link isn't nofollow'ed, etc.

I won't argue with your general point - you're right that "great" content isn't enough by itself. If you build the most beautiful house that ever existed on an island no one visits, you'll never win any architectural awards.

That said, I don't think there are many verticals where paid links are essential. In fact, I think chasing your competitor's tactics is often a good way to shoot for 2nd. In many cases, people are ranking in spite of low-value tactics, and finding the tactics the competition isn't targeting can give you a lot more leverage.

It's absolutely true, though, that even content marketing has to be marketed. I think you have to look outside of SEO. When I had content marketing successes on new sites, it came from relationship building. I pounded the virtual pavement (whether it be blogs, forums, social, etc.) - NOT for links, but to build relationships. I brought the eyeballs in, and when that hit critical mass, the links started to come. Even better - the links KEPT coming with little or no effort. Some posts generate new links 2 years after I wrote them.

The worst part about low-value SEO tactics isn't the risk of a penalty - the worst part is that you have to keep doing it every day. You haven't built anything but what you scrounged for that day. Strong content marketing takes a lot more up-front - no question - but it lasts and it builds on itself.

It looks like you've got both the "www" and non-www versions being indexed by Google, so you need to solve that problem first. You could either add canonical tags for each page or 301-redirect one version to the other. These sitewide duplicates are going to cause you problems down the road.

I think you also may be overdoing it with putting the entire 5-word name of your site at the beginning of every TITLE tag. I'd switch that to the end (except for the home-page). Put the most unique keywords up front.

Solve the www vs. non-www issue first, though. That could run you into a bunch of weird problems going forward and dilute your ranking ability.

Some links may be unremovable. If they're all coming from one domain, they'll likely just be devalued - Google is going to see this as a single pile of links (not 900 different entities) and probably ignore it. I'd leave it, in most cases, and move forward with higher-quality links. Blogspot isn't paid or a link network, so those links are just low quality.