Welcome to the Q&A Forum

It looks like for that second link, there is no text content on the page - only one big image with the specifications of the item. That brings the unique text content on the page down to almost 0. The picture of the man clutching his throat is also just a page with one image on it, and no text. Since neither the Moz Crawl Test tool nor Googlebot can see pictures, they are left with the impression that these pages are almost identical.

It's important to remember that the canonical tag is only one way to address duplicate content concerns - "duplicate content" doesn't always mean "canonical tag." Sometimes, when pages have very little content, they appear to be identical or near-identical even if their intents are different. In that case, I recommend seeing what you could do to make the pages more unique. With image pages, I wouldn't worry too much about adding content, but for pages like http://medresourcesupply.com/index.php?src=gendocs&ref=detailed_specfications &category=Main, where there is content, but it's contained within an image, I would recommend recreating the text contained in that image, but in HTML text. You could also expand upon the specs to make them more detailed, which will add to the unique content of the page and help users figure out if this is the item they really want.

I don't want to advise you to ignore Matt Cutts' advice but honestly in the same situation I'd probably just leave them be. I think Chris' suggestion of making sure you've got authorship set up is a good one. If you want to nofollow your links going forward you should do so, but there are probably better things you could do with your time than go back and change the existing links - especially if your inbound link profile is already pretty diverse and on-topic.

Hi Ken,

Thanks for looking to our blog as an example - I wish we could be the perfect example of on-page SEO that we want to be, but like a lot of big, oldish websites we've got a lot of room for improvement. Some of the things you've called out are on my wish list to fix. In general when it comes to SEOmoz.org, a "do as we say, not as we do" attitude can be useful. I'll try to answer your questions with "what we should be doing" instead of "what we are doing."

1. The only page within the blog that we might wish to self-canonical (as we do on individual post pages) would be www.seomoz.org/blog, which we probably should be doing.  I don't think there's as much point in self-canonicaling subsequent pagination, since as you say the blog pages are always changing. In general we do like to put a canonical tag on each page canonicaling it back to itself, so query parameters and other referring data don't cause problems with duplicate URLs/content.
2. Rather than noindexing paginated pages in the main blog, I would like to see us start using rel=prev and rel=next.  You can learn more about that here.http://googlewebmastercentral.blogspot.com/2011/09/pagination-with-relnext-and-relprev.html. The reason that we noindex category pages' pagination is that it becomes duplicate content, since all of those items are also appearing in the main blog's pagination as well.  We do want the main category pages to be indexed because we want people searching on those specific categories to be able to find them; we've had a lot of discussion in the MozPlex lately about how to give those pages some relevant unique content and make them more of a useful destination for people searching on specific topics. You may see some changes to these, although right now it's a bit low on my list of things to update.
3. We noindex the "show # of pages" pages because of duplicate content concerns, and I suggest you do the same.

Thanks for your question, and please let me know if you have any more.

Hi Scott,

1. That looks good to me! An additional factor you'd want to consider is how you're treating visits from the subdomains to the main site, and vice versa, in analytics - do you want those to be treated as referrals, or as part of the same session? - and configure accordingly.

2. If you've marked the subdomains "noindex, follow" Google will likely pass some link juice, but as usual in Moz Q&A the answer is "it depends" In this case, it depends on whether or not Google crawls the pages on the subdomain in the first place, and how closely-related Google perceives the subdomains to be to the main domain. So the answer is "some, probably, but probably not as much as links from unrelated sites that aren't noindexed."

3. From your question, it sounds like you're pretty familiar with the subdomains-vs-subfolders conversation in SEO, so I won't go into it here. Again, you're going to want to be really intentional when it comes to tracking on these sites to make sure you're properly tracking traffic between them.

This sounds like it could make a really interesting blog post once you've got it all set up!

In a post-Penguin world, I'd be leery of this kind of automated anchor-text-rich link building.  If you continue with this approach, make sure you're also getting plenty (i.e. a majority) of links without keyword-rich anchor text.  Otherwise, yes, they will both count as they are to different pages.

I think what your developer is talking about, and what Facebook does, is the idea of all of your content being on one page. Progressive loading or "infinite scroll" is when you scroll down to the bottom of a page (like e.g. a category page on your blog) and more content loads on the page itself, as opposed to having to click to page 2 of results to view more content.

The problem with doing this is that even though the content continues to load on the same URL, it's being pulled from another place - so anything beyond that first set of content is being loaded by a JavaScript call. That means that search engines can't index that content that's being loaded as the user scrolls down - and users with JavaScript turned off also won't be able to view the rest of your content. This can be a big problem on main product level pages like your client is thinking of doing, since any links to other products that are beyond that initial page load won't be crawled by search engines.

If you're going to go the infinite scrolling/progressive loading route, make sure that when JavaScript is disabled, there's a crawlable "next page" link to a new, static URL for the next page of results. Basically, make sure that there's a more old-school "previous page/next page" environment with static page URLs that search engines and users without JavaScript can browse, in addition to your progressive loading page.

Here's a link to a similar question from last year that has more information: http://www.seomoz.org/q/infinite-scrolling-vs-pagination-on-an-ecommerce-site

It's important to remember that the on-page grader only looks at how well optimized your page is for that keyword, which is just one factor out of many different things Google looks at to rank your site. Fixing technical problems and building additional links to your page and site will be very important to your ranking well. Good luck!

I would be cautious with doing too much article marketing - you're at risk that search engines will see your article on multiple sites, deem it duplicate content, and then decide your article on a submission site should rank instead of the version on your own site.  With the Penguin release, Google is also taking a closer look at links from article submission sites - I wouldn't be surprised if those links become lower and lower value as time goes by.

Definitely let Google discover your backlinks on its own - grow your link profile too quickly and they'll take a close look at your site to see if any of those links look suspicious.

There are plenty of more valuable ways to grow your link profile.  Mike King recently had a post on this topic at http://www.seomoz.org/blog/the-noob-guide-to-link-building - I suggest you check it out for new ideas on how to build links to your site.

Hi Jon,

I'll be doing a Mozinar on June 27th telling the whole story, and I'll have a big blog post with lots of juicy data in it to follow up. One reason we are waiting a bit is that there's a lot of noise in the signal right after a site migration, and I want to give you guys some realistic data.

The short answer is: So far it's gone very, very well - more smoothly than any rebrand/site migration I've ever worked on. We have lost a bit of (mostly long-tail) organic traffic, but less than I was forecasting. We've seen an uptick in bounce rate, but since we've also changed our funnel that's also to be expected. You can sign up for the Mozinar at http://moz.com/webinars, or view the recording later if you don't have time to attend.

Hi Christian,

What you're seeing is exactly what Google recommends for infinite scroll in the resource you link to. It breaks the page up into component resources (separate URLs) each of which could be accessed on its own. Their examples use dynamic parameters to break up into e.g. page=2, but if your infinite- or long-scrolling page isn't paginated content, there's no reason why each component couldn't have its own URL that is accessed as you scroll down.

I actually really like this method as a compromise between the "one long page with all the information on it" approach to web design and the "landing pages for people looking for specific bits of information" approach to SEO. For example, I often have SAAS clients who want all the information about what their product does to be one one long page. This is great for people who want to research the whole product at once, but makes it hard for me to optimize for keywords pertaining to individual features of the product. The solution is to have separate landing pages that talk about specific features, all linked together in one "product" page that scrolls using the methodology outlined in the Google resource you linked to. Plus, it means that people who are just looking for that one feature arrive on a page that's about that feature, instead of having to scroll to find what they're looking for.

With the infinite scroll situation, Google is only usually going to crawl and index what is available to the user before more of the page loads - so if you want Google to crawl and index all of the content on your infinite-scroll page, this is the way to do it. It's also better for users who don't have JavaScript enabled. I hope that makes sense and let me know if you have more questions!

Nofollow links do not pass link juice - Google may still crawl them for informational purposes, but they won't count toward your overall domain profile quality. I would only pursue sponsored posts/editorial links if you're confident that the blog's audience is a good match for you - i.e. treat it like an ad, not like link building.

I would be extremely cautious about buying sponsored links.  If the person doesn't disclose that they've been compensated for a blog post, they could be in violation of FTC guidelines.  A reputable sponsored post SHOULD nofollow their links, otherwise you would be in violation of Google's paid links policy: http://support.google.com/webmasters/bin/answer.py?hl=en&answer=66736

You've stumbled into a test we were running! Yes, putting it in the is the methodology recommended by Google - we were testing to see if having it in the anchor would work as well (looks like no, it doesn't). We're often running tests of this type so just because you see us doing something, doesn't necessarily mean it's the 100% best way to do it! You're better off reading the blog and Learn SEO sections for best practices information. Good catch!

The links to metro-manhattan may be losing a small amount of value as they pass through the redirects. Google has previously stated that this is not the case, but I think it's more accurate to say that it's not always the case, and I have certainly seen a measurable decrease in link value (usually around 15%) when passing through a redirect. I doubt that Google is confused about the two domains, though, since metro-manhattan has been redirecting to nyc-officespace-leader for 4 years and, unless I'm reading your question wrong, has never had content on it.

If metro-manhattan has more inbound links pointing to it than nyc-officespace-leader, AND is a better reflection of your current brand, it may indeed be worth moving the domain. I would not move the site from one domain to the other just based on the reasons you've outlined above, though. Moving your site from one domain to the other is likely to negatively impact your rankings and traffic in the short term as Google gets used to it being in the new spot. This is entirely possible to overcome through marketing the new site, but will not be a quick fix to the problems you've outlined. So if you want to move the site, and commit to marketing that site, go ahead, but don't do it just to capture link value pointing to metro-manhattan.

If you do decide to move the site, my advice is always to make a domain change separate from making other major site changes in layout, content, etc. So you would either want to:

• Move the site as-is, do some proactive link building to the new domain/reach out to people who link to the old domain to get them to update their links, get a marketing and promotion plan in place, and then once your rankings and traffic have stabilized from the move, start making improvements, or:
• Make the improvements to the site you want to make now, and revisit moving the domain after those changes have been in place for a while.

Your other option would be to reach out to sites that link to metro-manhattan and ask them to update their links to point to nyc-officespace-leader. That would allow you to get more value from those links, but I know it's not always possible to do. So like I said, unless you have additional compelling reasons to move the site beyond what you've said in your question, I would leave it where it is and focus on improving it.

I would add that one criterion we look at for the directories list is whether or not real people are actually using them. I don't think directories are a good tactic (and certainly shouldn't be your only tactic) for getting links, but can be a way to build brand recognition and drive some additional referral traffic, which is part of the reason why we don't just abandon the directories page completely.

I will say that we have plans to go a different direction with the directories page/list in the near future, although all of the specifics haven't been nailed down yet. It's definitely a resource whose usefulness we've questioned and we're working to make it more useful.

Does the page /blog/recipes/recipe-search have anything on it? With search results pages, I often recommend that people use rel=canonical to point back to the main search page. You should be able to change the tag search results page template in your CMS to automatically add a canonical tag that always points back to the main search results page. This will mean your search results pages won't be indexed, so this may not be a perfect solution if you're currently getting organic traffic to those pages. Google is known not to really want "search results in the search results," though, so it could be worth doing.

Another solution would be to go into Webmaster Tools/Search Console and tell Google to ignore the paramater ?akID%5B21%5D%5BatSelectOptionID%5D%5B%5D. Again, you should only do this if you're sure you don't want these pages to appear in the index.

Does that make sense? Let me know if you have any questions.

I agree that it doesn't make a lot of sense for Google to require the image itemprop - it's not the only example of Google requiring an itemprop that doesn't really make sense (for example, OperatingSystem is required for the WebApplication property, which doesn't make sense since it's...you know...a web application).

For what it's worth, even though the documentation says it's required, BreadcrumbList code that doesn't include the image attribute validates just fine in the Structured Data Testing Tool, and pages without the image itemprop in place are still getting breadcrumb snippets in the SERPs. So it may not be as "required" as Google says it is.

I think it's just something we neglected to add to our new robots.txt file! I'll see what I can do, but our Inbound Engineering team is pretty busy planning and building new awesome features so it may take a bit.

Matt is correct - building pages in a way that adds value to the user is the way to go. Building a separate page for each variation of a keyword is not only unnecessary, it can lead to lower rankings overall since Google has stated that these are not the kinds of sites they want to rank.

I would recommend very strongly against scraping content from other sites, or posting the same content on mulitple sites that you own, as well. Spend that time and effort on creating unique, expert, trustworthy content that adds value for your audience instead.

I haven't read the book you mention, but it might be time to find a different resource to learn about content marketing.

Adding a meta noindex tag can mean it takes a few weeks for a page to fall out of the index. These pages probably aren't doing you much harm, so if you wanted to just wait for them to fall out, that's probably fine (although I would update the tag content to "noindex, follow" to help Google crawl to the other noindexed pages). If you really want them out of the index faster, you could use the "Remove URLs" function under Google Index in Google Search Console, which will temporarily remove them from the index while Google is registering the noindex tags, or you can use the Fetch + Render tool and then Submit URLs in Google Search Console, which will cause Google to come back and crawl your pages and find the noindex tag.

Hi David,

The proliferation of links throughout the sites you left comments on is probably just a function of some weird trackback settings on their part - not a whole lot to be done besides disavow the links, which you've already done, although you might also try to delete the link from the original comment and see if that helps.

Your site sinking from 4th to 7th is not an indication of a penalty - if you'd been penalized, you wouldn't be anywhere near the first page (although I do think you are describing link building behavior that puts you at risk for a penalty - more on that in a minute). Instead, what I suspect happened is that the sudden influx of low-quality links has diluted your page's authority. Since a lower percentage of the total links pointing to the page are high quality, you have gone from sending the signal "great links, with some not-so-great ones too" to sending the signal "LOTS of not-so-great links, a few good ones" which sends a lower quality signal to Google.

Your question has a couple of big link building red flags in it, and I think it's likely you're at risk for a manual link penalty or a Penguin smackdown. Blog comments and directory links are classic examples of low-quality/spammy link sources. You have recently seen how blog comment links can result in a deluge of link spam to your site. Basically, anywhere that you can put a link on a site that you don't own, without a human person approving and curating the link, is going to be a link source that, at best, isn't valuable; at worst, it could actively work against you. I know that it's tempting to go out and grab these easy-to-get links, and sometimes it seems like it could be OK if the site is topically related to yours, but these are tactics that have been heavily abused by spammers and are beyond useless in modern SEO.

Instead, focus on getting higher-quality editorial links to your content. It takes a lot more work but is a much lower-risk strategy in the long term. These links have the added bonus of potentially being a source of converting referral traffic to your site as well. Here are some resources to get you started:

• https://moz.com/beginners-guide-to-link-building
• https://moz.com/blog/the-noob-guide-to-link-building
• https://moz.com/blog/targeted-link-building-in-2016
• https://moz.com/blog/case-study-how-we-gained-more-than-100-links-for-a-travel-website-via-content-marketing

Good luck!