Welcome to the Q&A Forum

Can you clarify - did you change them and then change them back? I'm not sure if we're dealing with one set of 301s or two sets, sequentially over time. Changing and changing back can get dicey.

Another option would be to use canonical tags - sometimes, if 301s aren't working (especially if you've changed them a couple of times), switching signals can help. I think Google starts ignoring certain signals at times, especially if things are changing too often. Using another signal (like rel-canonical instead of or in addition to 301s) can kick them back into gear.

You can use rel-canonical on other, non-blog sites. It does help to have a one-to-one relationship of pages between the two sites.

In theory, the 301-redirect is the better long-term approach, but in practice, rel-canonical seems faster and a bit more powerful in the short-term. So, David's basically saying to use rel-canonical for that initial boost and then do 301-redirects later to make it all permanent. It is a more technically complex approach.

It's really tough to predict. I think Eyepaq covered the basics well - you'll consolidate your link-juice, but you may harm your UK-specific ranking slightly. Whether the consolidation offsets the loss really depends a lot on your link profile, how Google treats your UK site (sometimes, English content, even across countries, is seen as partially duplicated), and your market focus.

If 80%+ of your market is US-based, you're not ranking that well in the UK, and you don't have the resources to really push two domains, my gut reaction would be to favor consolidation. If you have two separate marketing efforts int he two countries and half or more of your sales are UK-based, then you'd be taking a real risk.

I would check out the newish rel="alternate" hreflang="..." option:

http://support.google.com/webmasters/bin/answer.py?hl=en&answer=189077

It's specifically for helping same-language content rank correctly across regions, and Google seems to be pushing it publicly.

I recognize you're trying to be helpful, but I have to warn that, in this particular case, I think that's a bad idea. It would basically collapse all the products down into one page. This isn't a situation where the colors are variants of the same product (like 3 colors of the same t-shirt). The swatches ARE the product, in a sense.

They're not really true duplicates - we're probably detecting them as duplicates because there's so little on-page content. Ideally, you would get some kind of description in place, even if you just write them a few at a time over weeks or months. On the bright side, you do have unique titles. I'd at least populate the ALT text of the images, even if you do it from the data (with the name of the swatch). Anything unique content you can add will help.

Given the scope and that's important that these show up as individual products, you can't really canonicalize them without losing their ability to rank. It's always a trade-off, but I think it's best to let them rank and try to add some unique content and tags over time.

Pagination is a tough issue, and there's no one right answer (Google can't even agree with itself). If you're not experiencing ranking problems, then rel=prev/next is probably a good solution. It's a "soft" signal, IMO, and will help Google sort things out.

If you begin to have trouble, I tend to use META NOINDEX,FOLLOW on pages 2, 3, etc. In the vast majority of cases, these paginated results have no SEO value and no inbound links, so they add no value to the index. It really depends a lot on the scope of the problem.

I have to disagree on the risk assessment - true duplicates, on a massive scale, can cause something as severe as a penalty. It's not a Capital-P penalty in the traditional sense, but people who were hit by Panda due to massive duplication might as well have an algorithmic penalty. I'm dealing with client cases right now where duplicates caused much, much more than minor issues (80%+ traffic losses over months).

You've got the self-referencing canonical tags in place, which is about the best fix - unfortunately, there's no way to "undo" a bad canonical other than put a good canonical in its place. Hopefully, Google only picked up a few and, given the popularity of your site, they'll re-index pretty quickly.

Did you fix it? I'm only seeing one canonical now. That would definitely be bad - if Google picked up the canonical to the home-page, you could collapse a ton of pages into one and effectively knock them out of the index (and, by extension, any ability to rank).

I'll give you that - it's true that the result highlighting alone doesn't prove Google is viewing "shop ipad" as 2 words for ranking purposes. The tough part is separating out why these pages rank - I've seen plenty of URLs with underscores rank perfectly well, but there's no way to prove it isn't on other factors.

I think we're generally agreeing, though, that the potential risks probably outweigh what would probably be a very small boost at best. I'm also a little hesitant on having some pages with dashes and some with underscores, but that's probably just the UX side of me - it feels messy.

This is a great debate and some good comments all around, but I do think it's important to keep in mind that Matt's statements were from 2005 and 2009, and even the 2009 wording was a bit more nuanced. I strongly suspect Google has eased up on this early technical constraint. For example, try a search for "shop ipad" - you'll see this Apple.com URL in the Top 10:

store.apple.com/us/browse/home/shop_ipad/family/ipad/select

Google seems to have no trouble bolding "shop_ipad" as two words.

I still completely agree that, if you were starting fresh, I'd go with hyphens. A sitewide 301-redirect does carry some risk, though, so I think it boils down to this - are your pages ranking now for terms in the URL? If they seem to be ranking and it's just a matter of ranking better, I doubt the change would help much. If they seem to not be ranking at all and you suspect the URL terms are to blame, that's a different story.

Even in a perfect world, keywords in the URL are just one, relatively small factor. I'd want to see the data, but my gut reaction is to leave it alone until you need to make a sitewide URL change for other reasons.

It's tricky - these are http://googlewebmastercentral.blogspot.com/2011/09/pagination-with-relnext-and-relprev.html

To steal their example, page 2 of search results is going to need 2 tags, like this:

If there's no page refresh and you're using AJAX, it's possible your paginated search isn't being indexed at all. It depends on what you're showing to bots. Unfortunately, that's nearly impossible to tell without seeing the site.

I 100% agree with Marisa. I don't fully understand what you're trying to accomplish, but it looks like a recipe for disaster, to be blunt. You're going to moving link-juice around in really weird ways and basically splitting your efforts.

Why can't you house everything on "www.abcshoes.com"? Is there a technical reason you need the "...onlinestore.com" domain (such as a 3rd-party storefront)? I feel like we're missing a piece of the puzzle here.

Technically, rel=prev/next is more appropriate, but it can be really tough to implement and Bing doesn't honor it.

If the paginated search pages don't have inbound links, you could just use META NOINDEX,FOLLOW on them (pages 2, 3, etc.). It's a lot easier to implement and is still very effective.

It is a little tricky, but I think a link from a paid listing is still a paid link, if you want to push the letter of the law. Since you're not worried about the outbound link-juice, I agree with Keith that you're safer just to nofollow.

I share Alan's hesitation - it could look like cloaking, especially if a bot is making the call. If the pages aren't indexed yet, you could just "nofollow" the links - it sends the same signal transparently.

Home Depot is probably pulling it off with the AJAX/JS implementation, which is a bit harder for Google to parse. They also have a massive authority and link profile, so they can always squeak the small stuff by. You might not be so lucky. In general, it's best to stick to the standard practices and not get too tricky.

I actually find NOINDEX, FOLLOW a bit easier to implement and very effective in most cases. If you use rel-canonical, Google would prefer that you canonical to a "View All" version of the page (not to page 1 of search results). Practically, though, either will work.

I've also had decent luck with rel=prev/next, although implementation is very tricky. One other problem is that Bing doesn't honor it.

This subject can get very tricky, and no one (even Google's own reps) seem to have the one "right" answer. I find, in practice, that it depends a lot on the site and scope. Adam Audette has an article that shows just how complex pagination can get:

http://searchengineland.com/five-step-strategy-for-solving-seo-pagination-problems-95494

Are you securing a lead-gen form? One thing to test first - have you done an A/B test of the in-page form vs. a large call-to-action button? I've seen testing go both ways. Sometimes, the in-page form drives more leads, but sometimes a clear call-to-action to a separate form is better (really seems to depend on the form and the industry). If the call-to-action works just as well, it would be much easier to secure the stand-along form than your home-page.

Going secure full-site would help solve some problems, although you do need full-site 301-redirects then, and you've got to make sure your servers can handle it (https: needs additional overhead). It's really tough to tell without knowing the nature/scope of the site.

I'd be happy to close the private question, but if you want to dive into details there, it might not be a bad idea. Seeing the site would really help, since this is a tricky issue.

Got it - saw your other question. This is definitely a bit tricky.

The problem with "myths" like keyword count/density is that they were probably true once. In the last 90s (pre-Google), if you knew the exact density of keywords to use on any given day, you'd rank pretty well on the early search engines. It was a cat-and-mouse game. Things have changed a lot since then, though.

Yes, exactly - if you see a blog post with 100 comments and they're all using exact-match anchor text to 10 different interests ("Free Casino", "Cheap Viagra", etc.) it's a pretty clear sign that blog's been spammed to death. Admittedly, it's not always so obvious, but you can usually spot quality problems pretty quickly.

You can write canonical tags dynamically, but you need to canonical to the product specific page. Obviously, don't canonical every details.php page down to one. It could look something like this:

$prodDesc = "games-playstation-vita";

$prodCode = "B0054QAS";

$prodURL = "http://www.mydomain.com/".prodDesc."/".prodCode;

?>

I assume that the product description and code are generated from a database, so they should be available somehow to the header.

This is really tough to tell from generic examples, because the first thing you need to get at is the root of why these different URLs are being crawled in the first place. If "B0054QAS" is a unique product, why is it reachable by "playstation-vita", "playstation-vita-psp", etc.? That architectural problem has to be solved first, or any canonicalization is just a band-aid.

I generally agree with Alan (although I think NOINDEX, FOLLOW is ok, since these pages are unlikely to have external/inbound links), but there's no perfect solution for these types of pages. They aren't exact duplicates, but they may look low value to search. Given our current tools, canonical may be your best choice.

If you're talking about a couple-dozen pages, it's no big deal, and you could leave them alone. If the different filters are spinning out 100s of variants, then I would control them somehow.

I couldn't tell you just from those numbers - it's a lot more complicated than that. How many links are on the page, for example? Where will the link appear (in context, footer, sidebar)? Can you control the anchor text? Is the blog relevant? Does the blog have other, spammy links? Rand has a great post on the value of a link here:

http://www.seomoz.org/blog/10-illustrations-on-search-engines-valuation-of-links

All else being equal, in those two situations, I'd say a DA40 is a big jump over DA20. Toolbar PR3 vs. PR2 means almost nothing, and could be months out of date. The 150 Facebook fans could indicate that (1) has momentum - maybe it's growing while (2) is old/stagnating. It gets complicated fast.

I'm actually not sure I agree. From a theoretical, PageRank-passing perspective, sitewide links are better. From a penalty/risk perspective, 1000s of sitewide links can lead to a ton of links coming from very few unique domains, which can start to look suspicious. I actually think you might see less devaluation by limiting the footer links to a couple of strong pages on each client site.

Practically, I think Julie is right, but I have seen heavy devaluation of these footer links in the past year or two. They'll still count for something, but not a lot. The only warning I'd add is that I wouldn't create a situation where these are your ONLY links. You could risk looking like a link farm and even a potential penalty at that point. These easy links should be only one part of your link-building strategy.

I'd also highly encourage diversity. Mix up the anchor text, as long as it's relevant, and maybe even put the link different places. If you can get contextual links somehow (not footers or sidebars), that's a huge plus. The more you can mix it up, the better.

I think we say this mostly because it's really rare for a site to have all high-value links (as @authoritysitebuilder said). I'd agree with you that #1 beats #2 - it's just that that rarely occurs.

Once you've got #1, though, it doesn't hurt to add 5 easy links - relevant blog comments, guest post on low DA sites, etc. (not spammy, per se, just not as strong). Once you have a solid base of strong links, those weaker links will benefit you. Getting those extra, weaker links, is pretty easy at that point, so you might as well go for it.

The biggest danger, and where we really push diversity, is to have nothing but low-quality links.

(1) How many 301s are you talking about? As Highland said, you could lose link-juice (especially over time), since the new home-page for the blog won't seem as relevant to the redirected pages. I've heard of issues with 301'ing 1000s of pages, but those wer eusually temporary.

(2) Sometimes, post-Caffeine, Google crawls so fast that a 24-hour outage can cause some problems, but they're almost always temporary. You won't be penalized, per se - you could just see some rankings bounce while Google sorts things out.

I don't there's a set number or even set density these days - we just try to set some rules of thumb. In some contexts, you could have a keyword 20 times and it could look natural. In others, 10 might look odd.

Typically, I think certain tags, like TITLE, may be looked at in isolation. Even 2-3 repetitions in a TITLE tag can look spammy, especially to users. It's rare that would get you in trouble with Google (they'd need to see a pattern across the site, IMO), but the extra repetitions don't usually have any value, and may actually decrease the ranking value of other, unique keywords.

In general, focus on variations and natural language. Don't get hung up on word counts.

Just a note that it may not impact MozTrust. mT is based on links from a trusted set of seed sites, so unless it's a link from one of those sites, the mT wouldn't change. mT doesn't measure your total authority - mR and DA/PA do that.

301 is probably best here. There are two tricks here, though, and it can get messy:

(1) You'll need to link to the HTTPS version in your internal links as well.

(2) If you use relative links (like "/about.aspx"), then all the navigation links from your secure home-page will cause Google to crawl the rest of the site with HTTPS, possibly creating mass dupe content.

Practically, there's a fair amount of risk in just securing your home-page, and it goes beyond the home-page itself. I'd proceed with caution and really evaluate the pros and cons.

Present day, Google has no issue with crawling/indexing secure pages, but Naghirniac is right - they will be seen as duplicates. The canonical tag is probably your easiest solution.

The best canonical solution, though, is to use the same URL consistently. Out of curiosity, why are you building links to the secure versions?

I haven't seen any evidence that it's a problem, but John's point is correct - Bing does officially say not to do it. Actually, Google originally said this, too, but then eased off (if I recall correctly). It's gotten so common that I don't think either engine can really penalize it, honestly. I do it all the time.

I do understand - when you don't control the budget AND your competitors seem to be getting away with every trick in the book, it is frustrating. It's tough when you believe white-hat really works and also realize Google rewards too much bad behavior at the same time.