Welcome to the Q&A Forum

I haven't found standard shipping info text to be much of a problem in terms of duplicate content on product pages, but if you're concerned about it I would consider an iframe.

While hiding the content in a non-crawlable .js file might work, it isn't advisable. First, Google PreviewBot can parse javascript on the page and see the content in many cases, as is necessary to generate the site preview and visual cache of a page. Second, anything that keeps Google from being able to properly render the page may be considered suspect.

Please view this video by Matt Cutts: http://www.youtube.com/watch?v=B9BWbruCiDc

Essentially, he says "Don't block Google bot from css or javascript".

CFSSEO,

Can you share the site so we can have a look at it to provide a better, more specific, answer for you? If not, I hope this general answer will help:

You should only have one canonical page per product. Even if there are product variants (size, color, etc...) those should typically be handled with a drop-down selector or checkbox on the product page instead of creating an entire new product page for each variant.

If this is a large affiliate or drop-shipping website where you are pulling in feeds from different merchants/distributors/stores the bitter pill you may have to swallow is that such sites are dying out in the SERPs because Google doesn't think they offer anything above and beyond their own search results. Someone wants to buy a Blue Widget and they go to Google and type "Cheap blue widget" and a bunch of merchants'/distributors' stores show up... why do they need to put another site in-between the searcher and their destination? The only way to work within this paradigm is to prune the site down and focus mostly on category-level searches, price comparison searches, etc... and by offering more useful features, including robust customer reviews for products and merchants, tools to help the shopper compare options, in-depth buying guides to educate shoppers on their choices... but thousands of product pages with default manufacturer content will no longer do the trick, and may actually harm the rest of your site.

This may or may not apply to your situation. It is tough to give advise like this without knowing the site in question. However, I hope it has helped point you in the right direction.

Having too many 302s won't necessarily harm rankings in-and-of itself. However, don't expect the pagerank from those old pages to be applied to the new page / home page.

Also, redirects, whether 302 or 301, which link from a specific page to the home page are often treated as 404s anyway.

Ruben,

Woo Commerce for Wordpress is a pretty good solution for a small eCommerce site. If you're going for a bigger site with more robust inventory needs, including integration of a Point of Sale (POS) system you might go for another solution, such as BigCommerce, Volusion, Magento... which will be more expensive but also more scaleable.

I think Woo Commerce will work for you based on what you've stated so far.

I typically keep category names out of the product URL. It creates problems, especially when the product lives in multiple categories. You can use a rel canonical tag, but I'd rather just not have to deal with it. Here is how I do it:

www.domain.com/products/product-name/

www.domain.com/category/category-name/ and
www.domain.com/category/category-name/sub-category-name/

This convention has several advantages, including easy segmentation of the site to determine, for instance, how many product pages are indexed in the SERPs.

One could argue that having the category name in the product URL is good for SEO because of the keywords, but I would argue that putting the product farther down in the site structure, and the non-canonical URL issues related to certain taxonomies (e.g. multiple categories for a product) negates what little benefit keywords in the URL have these days.

Of course not all eCommerce platforms really allow the structure above. Magento, for example, will allow you to put products in the root, but not in the /product/ folder. The product will also be viewable on the category-version of the URLs, but they will have a rel canonical tag pointing to the root directory version.

Good luck.

Ltsmz,

You don't necessarily need to put the keyword "ProductName Review" several times on each page. However, if what you're showing on that page is indeed a review of that product, then I don't see why "ProductName Review" wouldn't be included in the copy, or header tag of the page. Once or twice should be plenty in most cases if you are repeating the words "ProductName" and "Review" each separately elsewhere on the page.

1. I would optimize for both, as outlined in my answer above.

2. Yes it makes a difference. Having the word "review" appear in a sitewide sidebar or footer is going to have much less of an impact than having it within the body of the content on the page.

3. Interview industry experts and ask what their favorite products are for a specific need. Email the winner of a comparison, possibly giving them a badge (e.g. "Editor's Pick") to display on their site. Create high-quality, useful comparison charts that compare the price and features of each product within that niche. Generally just have a really useful, well-designed, regularly updated website.

John,

The googleoff and googleon tags are meant for Google's enterprise site search product, Google Search Appliance. They "shouldn't" have any effect on the public index. Do you have an example where you can prove they work in Google search?

Hello Larry,

Yes this is a problem for you, and could be harming your rankings. Google has hundreds of those review pages indexed from your domain already: http://goo.gl/Ey7w4c

There are a number of ways to deal with this issue, as outlined below:

1. The quickest, easiest way to get them out of the index is to use the URL Removal Tool in Google Webmaster Tools to remove the entire /reviews/ directory from the index. Then you can place a disallow statement in your robots.txt file to block the entire /reviews/ directory from being crawled and indexed again in the future. The downside of this is that you won't get much, if any, "credit" for links pointing to those review pages. However, I doubt anyone is linking to them to begin with.

2. You could put a rel canonical tag on the review pages, which would point to the product page for which the review was left. For example, http://www.audiobooksonline.com/reviews/review.php/full/0743554337/0/name/desc would have this as a rel canonical tag:

3. You could 301 redirect the review pages to the product page, as someone else suggested, but I think this would cause all sorts of unintended errors and redirects on your reviews platform. I don't recommend this route.

4. You could use a noindex,follow meta tag on the review pages.

Googlebot and Mozbot, and their respective indexes, behave differently. You're not always going to get the exact same numbers from GWT that you get from Moz.com. Use Moz.com to identify critical issues on your site, rather than expecting it to give the same broken link count, 404 error, duplicate page count, etc... as Google.

I hope this helps give you some options. Also, I recently wrote an article that I think might help you:
eCommerce Product Review Solutions

Good luck!

Good to know dogids. Thanks for the update. Since my response above I wrote a post comparing eCommerce review vendors, including Revoo, on SEO best practices. Here's the link if you'd like to read it, but I thought Revoo had a pretty good system overall. They're definitely not like everyone else - http://www.seoverflow.com/ecommerce-product-reviews/

Takeshi gave some great advise, and I agree.

I do like the idea of varying the footer navigation by category though. For example, you might interlink Men's Shoes with Women's Shoes, but you wouldn't want to link to Women's Shoes from Men's Shirts. Keeping the footer from being site-wide and making the links more relevant to the page their on will, I think, allow those footer links to carry more weight than they would otherwise.

Whatever you choose, make sure it's best for your users. Do that and it's difficult to go wrong.

Steve,

Yes that would be cloaking. I wouldn't do that.

As Pete mentioned below, your only real options at this point are to make some of the content, or new content, available for public use. If you can't publish abstracts at least, then you'll have to invest in copywriting content that is legally available for the public to get traffic that way, and do your best to convert them into subscribers.

Amy,

I'm sorry you have to deal with a snafu like this.

I noticed that the content most relevant to the keyword on the page tends to be pushed down below what reads like "SEO copy". For example, on the page about moving from the UK to Adelaide, Australia the entire first screen's worth of content is general info about Adelaide, and you have to scroll down pretty far to get to anything specifically about the UK. Another example is on the country-level pages (e.g. /spain, /thailand) the "Top Cities" with their score-bar and relevant links are pushed down below several long paragraphs about the country. Perhaps your users are already aware of the basic information about the country if they want to move there, and the more visually-appealing and helpful "Top Cities" area should be moved above? If you're worried about the SEO ramifications of moving the text content down you could just try this out as a test on a few country landing pages.

Let's think about the visitor for a minute. If I searched for "Moving to Dallas from the UK" and landed on this page I would not be happy: http://www.movehub.com/usa/dallas/move-to-dallas-from-uk . I would be looking for information on how to move to Dallas from the UK, things like visa requirements, good Vs bad neighborhoods, cost of living, job market, etc... Instead, all I see on my screen is a lot of fluff copy about how Texas isn't all about oil production anymore, and sidebar links to content about international container shipping costs, and moving to cities in Australia and Canada. The site needs a more intelligent way to show relevant content and links on a per-page basis. Some examples: Instead of showing links to Adelaide, Brisbane, Perth, Sydney, Toronto... on that page, show the cities closest to Dallas, like San Antonio, Austin, Houston... Instead of showing the general "about Texas" type content at the top, move it down or get rid of it alltogether and show the content that was once hidden below the fold up top: Moving To Dallas from the UK, and Comparing Dallas Vs London. Put in some links to pages about things like visa requirements from the UK to the US, and some job-search assistance (top employers in Dallas?) and you'll have a much more useful page.

More specific to the problem you experienced, however, if the content is no longer on the page it may just take Google some time to recrawl all of the old URLs again and see the updated content without all of the "keyword stuffing" they may have misunderstood. My advise would be to refresh your XML sitmap with new lastmod dates and resubmit it to entice Google to recrawl the pages again and see that the excessive keyword use has been fixed.

Hello KempRugeLawGroup,

If the old domain had external links pointing to it that are relevant for another domain you may want to consider 301 redirecting it at the page level. It may be too late for that, however, if it has been parked for awhile.

Mass 301ing an entire domain to a single page on another domain (e.g. the home page) is just going to treat the old domain URLs like 404s instead of passing on any pagerank from external links to the old domain. You could do this for user experience, but it wouldn't help for SEO (probably wouldn't hurt though, unless the old domain had spammy backlinks or an existing penalty).

If the old site had a penalty, spammy backlinks, or no backlinks, I would just leave it parked to make things easier until you figure out what to do with it.

However, if it has good backlinks that could be sending visitors who are relevant to your new site to a "parked" page I would think about redirecting them to the new site.

I hope that helps and will leave this open for discussion from others in case there are differing opinions.

The data in GWT is not always updated regularly. Also, for a new site that has never been indexed before and has no, or few, external links, it would not be surprising to experience infrequent crawls. The more links you earn and the more of a history of fresh content and updated pages you develop, the more often and deeply you'll be crawled.

As Davenport-Tractor mentioned, an XML sitemap submitted to GWT will also help if you haven't done that already.

If most of your pages are indexed when you do a (site:yourdomain.com) search on Google I wouldn't worry about it too much. If they aren't indexed, you may have a problem, such as inadvertently blocking the crawlers via robots meta tag or robots.txt file. I'd have to see the site to know that though.

Hello Amelia,

Here's another Google piece about how to move to a new domain for rebranding:

https://support.google.com/webmasters/answer/83105?hl=en

Also don't forget to use the change of address form in Google Webmaster Tools, as discussed here.

You can use an HTTP status code checker to make sure the response codes are correct (e.g. 301 instead of 302).

Hello BeytzNet,

When people, including Matt Cutts, explain how the flow of pagerank works they tend to do it in simplistic terms to avoid confusing the situation, and to make sure that everyone understands the fundamental concepts.

Yes, at a fundamental level, the more links on a page the less pagerank flows through each of those links. However, it is quite a bit more complicated than this. Navigation links are treated differently than in-copy links. Footer links are treated differently than those at the top of the page. Sitewide links are treated differently than single links, etc...

When looked at this way, you can see how a link from within the body of a post is going to probably flow much more pagerank than one of the site-wide navigation links at the top of the page.

I agree with Maximillian that you should think about the best experience for your users. Here's an idea if you don't mind creating a totally separate page template and navigation...

Just show the top-level category page navigation on the blog instead of the complete drop-down list to sub-categories. This will drastically reduce the amount of links on the page while keeping the visual user-experience much the same for continuity and convenience. You can put the blog navigation in the sidebar.

That said, there really is nothing "wrong" with changing the look and feel of the blog from the main site if that is what you want to do. Just think about the visitor who has five items in their shopping cart already and decides to click on a link to a blog post, which then feels like a totally different website.

Federico brought up some good points when he suggested sub-directories above, but I'd go with sub-domains in this case. If a service provider can't figure out how to build their own website they are unlikely to get any decent links, let alone one from CNN. Policing them all would be expensive, possibly even unscaleable for the business model.

I think the use of subdomains as outlined in his original question would be one of the few proper reasons to have them.

Just make sure you verify each and every one of them in GWT so you know if/when any "manual actions" or important messages come through from Google.

Getting SEOs to agree on sub-domain Vs sub-directory Vs separate TLD can be like trying to herd cats sometimes. That's just the way it is.

That is a debatable question. In my opinion, I seriously doubt Google would have a problem figuring out that this is the same page. I don't think it would cause you to lose any pagerank.

However, if you just put a rel canonical tag on that page it should take care of it without even needing to worry about the redirect.

Here is how you have it set up now:

http://www.over50choices.co.uk/ (200 but browser removes trailing slash)
http://www.over50choices.co.uk (200 and should be your canonical URL.)
http://over50choices.co.uk/ (301 redirects to http://www.over50choices.co.uk/ )
http://over50choices.co.uk (301 redirects to http://www.over50choices.co.uk/ )

You should be fine, but if you wanted to change something I'd make the redirect from the non-www. version go to the version of www. without the trailing slash (i.e. http://www.over50choices.co.uk ) instead of http://www.over50choices.co.uk/.

You must have just entered over50choices.co.uk into the Moz tool without www., which is why it is letting you know that the tool decided to show for www. instead - though you have the option to force it to look at non-www if you want.

I wouldn't worry too much about it, other than choosing between www. with or without trailing slash.

%E2%80%9D is the UTF-8 encoding for the unicode character of the right double quotation marks (double quotes). This is a somewhat common issue with some Wordpress plugins. Please refer to the two posts below, especially if you are using either the Yoast's Google Analytics plug-in or a social plugin called Links Alpha.

http://moz.com/community/q/e2-80-9d-breaking-the-url-in-wordpress

http://moz.com/community/q/google-indexing-thousands-crazy-search-results-with-25253

As for the discrepancy, it could just be that Google's crawlers are a little "smarter" than Moz.com's in figuring out when unicode characters are used to represent double quote marks. No two crawlers are going to behave exactly the same.

You should probably set the www.website.com/the-mayflower/ version as the canonical. So the page on the subdomain, and all other copies, would ALL have a rel canonical tag that points to www.website.com/the-mayflower/.

I wouldn't block the subdomain based on what you've said, but complicated issues like this are difficult to fully diganose and prescribe fixes without seeing the site.

I've dealt with similar issues with robots.txt blocks of the entire site, as well as robots meta noindex tags. You should be fine now that you've taken the noindex tag off, and the old pages are redirecting. It may take longer for Google to update their index with the new URLs, but otherwise I don't think you need to worry too much. Maybe resubmit the sitemap and do another fetch on key pages.

Good luck!

On the sample page you provided it would be much better to label the photos with alt attributes like this:

Main image:Flatiron Loft for Rent

Hallway View for Listing #278
Window Views of City for Listing #278
Etc...

The point being to describe what the image is instead of worrying about what keywords to use.

What you're doing now probably wasn't the cause of a major traffic drop, but it probably isn't helping much either and there is a possibility that it could be seen as keyword stuffing.

From - http://www.amazon.com/gp/help/customer/display.html/?nodeId=508088

COPYRIGHT

All content included in or made available through any Amazon Service, such as text, graphics, logos, button icons, images, audio clips, digital downloads, and data compilations is the property of Amazon or its content suppliers and protected by United States and international copyright laws. The compilation of all content included in or made available through any Amazon Service is the exclusive property of Amazon and protected by U.S. and international copyright laws.

In short, you don't own those reviews even though they are about your product. The reviews were left on Amazon.com, and are therefore the property of Amazon.com.

Hello BarrelRoll42,

You should easily be able to find out if Google is indexing them by doing a site:yourdomain.com search on Google. But to answer your question, it sounds like you should probably delete them and let them 404. If Google HAS indexed them you may also need to use the URL Removal Tool in Google Webmaster Tools.

One last thing. Please do start a thread for your own question next time, as we try to keep it to one question per thread.

Thanks!

They're still relevant, especially for image search. I'd change one thing in your strategy though...

From "focused on primary page keyword"

to

"focused on describing the image"

Assuming you're using images relevant to the topic you can do both, but when you have to make a choice between those two choose the latter.

I'm glad to hear you've been sorted out Lawrence Neal. I find it interesting the the other Lawrence saw something similar, and I'll ask around to see if it was a glitch that other people have noticed too.

For anyone reading this wondering what Mr. Neal was referring to in regard to rel canonical / href lang conflict, there's a good writeup of it over at Dejanseo.com and Gianluca Fiorelli mentions it in his comment on Dr. Pete's Rel Canonical uber post here on Moz.

Also see answers on Wesley's other question thread here:
Should we noindex 65,000 product pages all at once or in steps? (paraphrasing)

I think you would be fine doing it all at once. The gradual roll-out thing Matt Cutts was referring to had to do with adding pages, not removing them.

You could also use the URL removal tool to get them taken out of the index faster once you have added the noindex,follow tag. You can remove an entire directory that way in one fell swoop as long as you don't need any other pages in that directory to be indexed either.

This is a smart move for a lot of eCommerce sites that like to drop ship products and use manufacturer or distributor-supplied product descriptions at a level that is not scaleable for rewriting on their own. In such cases I optimize the best performing products, get the rest out of the index, and rely heavily on category pages for bringing in search traffic. It isn't the best situation, but it's better than keeping a Panda penalty.

Luke if you're writing these manually I doubt you could possibly write and publish them fast enough for it to be any problem at all. If this is automated and you're going to be releasing thousands, or tens-of-thousands... of new title tags I'd slow down and do it in segments, not for fear of any "penalty" but because it's always good to test things out before making changes that big unless you know for sure it is a change for the better - which it sounds like it is.

Your question was: "Any idea how my site can rank on google?"

The short answer is: Write unique content, especially on product and category pages, and get quality links.

The longer answer is:

Write unique, useful content for every product page on your site. Yes, you can outsource this but go for the highest quality offered, not the cheap stuff. Try odesk, text brokers, elance, helium, writer access, etc... and go with whichever one you like best.

Start EARNING (as opposed to "building") backlinks by providing a great user experience with good, useful content. This is hard to do if you "don't want to write any article". Something as simple as outlining the reasons why it isn't a good idea to use an iPhone 4 case for your iPhone 5 might be a good place to start for the example above. Then you can interlink to your store's category pages too. The promotions you're doing are a great link-earning opportunity. Be sure to let bloggers (those in the money-saving / deal niche especially) know about it in a helpful, non-intrusive way.

Oleg makes some good points about putting those links on an index page for the articles (like the articles "section" of the site, which would be linked to from the home page) instead of putting every article link on the home page. However, I would argue that if these articles aren't helping "your website accomplish its goals" they shouldn't be on your site at all. Even if "SEO" is considered a goal of your website, you're bound to fail if you publish article-style content just for that reason.

So in your strategy I would really think about how this content is helping your users. For example, where in the process of their journey on your site they might need/want to see that content (e.g. 'How to Choose a Widget' might be linked to from your widget category, or 'Ten Ways to Use Your New Widget' might go on a widget product page)? That would be where you want to link to it from.

I always wonder though, how much non-link data is lost on such a page that could have an effect on rankings... such as the age of the page?

Bryan,

If I were removing 100 product pages from an eCommerce site because they barely convert I would approach it this way:

#1. Run the URls through a tool to see which ones have external backlinks.
Often times none of the pages will have any external backlinks, and those that do are usually not very good. If this is the case - or if you "really" aren't able to do any 301 redirects (and if so that's something that needs to be fixed) - skip to step #3. Otherwise...

#2. 301 redirect those with external backlinks to the most relevant page, be that a similar product or the category page directly above the product to be removed. Try to avoid redirecting them all to the homepage or some other "catch all" page, as these may be treated like a 404 by Google.

#3. Simply remove the pages and show the custom 404 page that suggests other products, or whatever messaging you want to show there (e.g. "This product has been removed from our catalog. Please see these other products...") and be sure to check the http header response code (lots of free tools for that) to ensure these URLs actually serve a 404 response (note: This should show up on the removed URL, as opposed to redirecting the visitor to another page like .../404.html).

#4. Since the now-removed URLs are not linked to from anywhere, either internally or externally, it could take awhile for Google to recrawl them and see the 404 error. If you need this to happen more quickly, such as when dealing with duplicate manufacturer descriptions and removal of page to recover from Panda, it may be wise to provide some type of html sitemap file listing out the URLs long enough for Google to recrawl them.

I would not block them in the robots.txt file, as that could result in Google not seeing the 404 and not removing it from the index (though they will cease to show the meta description).

Hello Coraltoes,

I would not rely on a nofollow attribute for those links. Instead, you could approach this in several different ways...

#1. Don't worry about it. If you're talking about an extra few links it really isn't going to make that big of a difference, though the text link should probably be first since it has the best anchor text.

#2. Remove them and rely on sub-navigation in the sidebar for sending people to the "featured" section of each category.

#3. Put the Featured fly-out navigation in javascript, while keeping the rest of the main navigation in HTML/CSS.

#4. Sometimes it is possible to wrap a single href tag around an image AND text, which would cut the superfluous links in half.

If you discuss these options with your developers perhaps some other ideas will come to light as well.

Good luck!

Hello David,

There are several great answers here, and I'd like to add one more option. The inbound marketing companies on the Moz Recommended List have all be vetted by Moz.com and should be reputable.

If I were going to hire an agency one of the least of my concerns would be which tools they used. Do your research to find out what kind of reputation they have in the industry and trust that they know what they're doing.

Good luck!

Hello Colin,

Since it sounds like you're not using any kind of CMS, and are coding an HTML/CSS site yourself, it will be a little more difficult to find an "app" to do this since most publicly available apps/plugins/widgets/modules are created for use in an existing CMS. There are some scripts out there built with JQuery that might work. Just Google "JQuery gallery" or "JQuery lightbox" to get started. If one doesn't have everything you want you should at least be able to get 90% of the way there, and can either tweak the code yourself or hire a developer to get the rest of the way there.

This one looks promising:
http://fancyapps.com/fancybox/

Also check out:
http://www.intelliance.fr/jquery/imagebox/
http://www.pirolab.it/pirobox/
http://www.no-margin-for-errors.com/projects/prettyPhoto-jquery-lightbox-clone/

It may also be possible for you to use one of the CMS-specific plugins with some tweaking. You could try out some of these on a free Wordpress site to test them. When you find the one you want you could try customizing the code yourself, or hiring a developer (having an existing plugin for him to see/use will save you money) to do it for you.

Good luck!

I like that idea in theory and in practice for some sites. But in practice for most of the sites I deal with I have enough trouble trying to get buy-off for custom copy on each product page, let alone five or ten totally unique, separate descriptions that don't all sound boilerplate.

I'd love to see a few examples though if you don't mind sharing. You can PM me if you like.

Cheers,

Everett

EGOL,

Great question / observation. I'm curious about this too, but do not know how a test could be designed that would rule out other factors. Glancing at the SERPs, I don't see any indication in terms of placement that this is a major factor since results #2, #3 #5-#7 don't have that data, but that isn't exactly scientific. On the other hand, those that do have it seem to be ranked in order of the total amount of items (i.e. 578 Amazon, 484 Zappos, 85 Shoebuy).

Regardless, nine times out of ten I recommend having a single product page with variables as something the user can choose from a drop down, as opposed to different pages. With this approach, in light of your question, it might be worth considering the option of having all of the variables listed as individual product pages on a category page, and using a rel canonical tag on the product pages to canonicalize them all to a single product page. This way you inflate your count on category pages, while avoiding the issues inherent with having separate pages for each variant.

If it's Amazon you are concerned with you could probably just leave them as-is. However, you could also run a scan of your site using a tool like Screaming Frog or OSE and look for outbound links going to Amazon.com and other known affiliate sites. This will at least make tracking them down to manually update them easier.

You could probably have this done at the database level too, though that is something for which you'll probably want to hire a DBA.

The following article has the best idea I could find for your situation, though it still isn't perfect because you could be applying a nofollow meta tag (as opposed to a nofollow link attribute) to pages that have both affiliate links AND editorial links that you want followed. At any rate, I hope this helps: http://www.bradlinder.net/2011/05/how-to-add-nofollow-attribute-to-all.html

Hello Andrew,

If you are on Wordpress here is an article explaining how to disable trackbacks/pings, which I advise doing for most sites. It also shares how to remove trackbacks from previous posts by running a SQL query on your database.

Alternatively, if you want to get the pagerank to pass through those links (inadvisable since 99% of them will be spammy, in my experience) you can change the redirect to a 301. See this Yoast article for more info on how to do that. It was written in 2007, but Joose de Valk is ahead of his time and I think it still works.

Every CMS is different so if you're not on WP please let us know what you're using. Of course you can always just block the /trackbacks/ directory from being indexed or crawled via the robots.txt file. I wouldn't do that with the /feeds/ directory though, as I find them useful.

I hate to bring up an old thread, but here's the answer if you're still looking for it:

http://yoast.com/google-wordpress-and-trackback-urls/

Cheers.

Can you provide one example of each?

The answer at this point is: It depends on what the order page looks like.

My gut instinct is that the best solution will be "something else". But it would help to see the pages so we can assess the value and uniqueness of each (i.e. are they both worthy of being in the SERPs; do visitors need to see them both; are they both 100% unique, etc...).

If all other things are equal AND you all have some crawlability issues, yes it could be a major factor. For example, if you have deep pages that are not indexed well then linking to them outside of the sitemap could help. Just because a page is in the sitemap doesn't mean Google wants to index it. If it has zero other links (either internally or externally) and the content is a little "thin" they may choose not to include it. Adding a decent internal link could make the difference there.

If all other things are equal AND those pages are indexed and linked to from your crawlable navigation, yes it could be a moderate factor if your competitors are interlinking from body copy and the only place you link from is the navigation - as these types of links are not treated exactly the same.

That's the short answer. The long answer is - it's complicated. Feel free to stop reading here if you really don't care about possible complications.

Some sites go too far on their interlinking strategy and end up with a penalty. I've seen this happen on a few Wordpress sites that use a plugin to automatically link every single instance of a phrase to a specific landing page. That just doesn't look natural, and you could end up getting filtered or penalized at the level of the page or, more likely, for that specific anchor text.

Another thing to look out for is future scaleability. For example, I don't often like to hard code links to specific products within the body copy of an eCommerce site because products tend to come and go. What you'd end up with a few years later is an avalanche of 404 errors that you'll have to manually go in and fix. I tend to leave product interlinking automated (e.g. Related products widget) and save the hard-coded links for linking to category pages, which tend to stick around longer.

I think this is actually a really good question. The main reason most SEOs these days don't "sculpt" or "shape" with nofollow links anymore has to do with the fact that they will still take away from the total amount of pagerank available to be passed on to other links on the page. So the question I'm reading above seems to be:

Do<a data-href...="" links="" still="" take="" a="" portion="" of="" pagerank="" away="" from="" the="" total="" pr="" available="" to="" be="" passed="" on="" other="" same="" page?<="" p=""></a>

<a data-href...="" links="" still="" take="" a="" portion="" of="" pagerank="" away="" from="" the="" total="" pr="" available="" to="" be="" passed="" on="" other="" same="" page?<="" p="">My answer is "I don't know" but I'd like to see a test if anyone can think of a way to try it out.</a>

<a data-href...="" links="" still="" take="" a="" portion="" of="" pagerank="" away="" from="" the="" total="" pr="" available="" to="" be="" passed="" on="" other="" same="" page?<="" p="">However, even if the test came back saying "No, these are treated differently and do not currently affect the total amount of PR available to other links on the page" I still would not use it for the purpose of pagerank sculpting. The reason is that how Google treats these links today can change tomorrow, making "tactics" like this a bad idea IMHO. It just leaves a mess for either you or some other poor SEO to cleanup later.

If I don't want pagerank to pass through a link on a page I simply don't put the link on the page. In extreme circumstances where there is no other way around it I might consider obfuscating the link with some javascript, for instance. However, even if you block the .js file that handles this "link" in the robots.txt file Google still executes it (as you can see when viewing the cached version on Google for pages that do this).</a>

Hello Henrique,

2. I mean putting /ads/ on a totally separate domain (e.g. youraffiliateprogram.com/ads/) which then redirects to yoursite.com. This way you can totally block that entire domain if you wanted to. When an affiliate links to youraffiliateprogram.com/ads/pageID the user who clicks on the link will get redirected to yoursite.com/pageID.

This set up has the advantage that you can block the entire affiliate program domain from being indexed in the robots.txt file if you wanted to, or you could try to use a 301 redirect and benefit from the links - until it stops working or you get penalized, at which point it you could block the affiliate program domain in the robots.txt, and/or change the redirects from 301 to a 302 rather quickly.

To answer your last question, yes you could simply change the redirect from a 301 to a 302 and that should solve the issue at the final landing page level, but since a lot of affiliates are still linking directly to your domain prior to getting redirected it could still cause issues. For example...

Affiliate A links to YourSite.com/ads/123 without using a rel nofollow tag in the link. This is still a link to your site from them, regardless of what happens next.

YourSite.com/ads/123 proceeds to 302 redirect the visitor to YourSite.com/123. The won't bass the pagerank on to your landing page, but it did nothing to stop the fact that you have a direct, followable affiliate link going to your site.

If you put in your Terms for affiliates that they have to use a nofollow tag in links to you, and you supply the nofollow tag in the code when they are "building" links from within your affiliate system, you should be ok. You may also want to block /ads/ in the robots.txt file just to be sure if you're going to take this route.

Hello Tesbat,

It sounds like you are using a content delivery network (CDN) to serve images, which is pretty standard. However, there are some best practices that I like to follow, and if your traffic from images is really that low it could be that you are already not following them.

First, I like to have a CNAME set up so the CDN exists on your own site's subdomain (e.g. cdn.yourdomain.com) as opposed to another server, (e.g. yourdomain.akamai.com). Every CDN company and host will differ on the details of how to set this up, but if you tell them you want the images to exist on a subdomain of yoursite.com I'm sure they can figure it out - if you don't already have it set up that way.

Second, don't change the filenames or basic folder structure. For example, if your images followed this format (/images/image1.jpg) and you went to a CDN that either renamed the images or put them in a totally different folder structure (e.g. /images/0000123.jpg or /CDN/assets/image1.jpg) you could lose some of the momentum you built up over time, especially if those image files are not redirected. Here's a case where that happened to someone.

Third, set up and verify the new CDN subdomain in Google and Bing Webmaster Tools. Add the geotargeting for the country you are targeting in case the IP address of the CDN is in another country. Typically this is not an issue for most sites, but is worth considering if the CDN is hosted in another country.

Fourth, continue to optimize your images with appropriate file names and alt attributes, and with the title attribute (as you do) when linking to larger versions of the image.

Don't worry about those links. I'd rather link to a bigger image than just have the thumbnail sitting there by itself. Just make sure you're actually linking to the image file, as opposed to another web page on which the larger image is embedded (e.g. image nodes in Drupal, or attachment pages in Wordpress).

Here are a few pages from Barry Schwartz worth reading:

• How Barry set up his Amazon S3 CDN
• Does Google Like CDNs?
• More from Google on CDNs and SEO

Hello Andrew,

It sounds like you're asking about how you should format your on-page factors to rank for searches in which people are typing in a specific year, or range of years. If this is the case, I think you will need to include each of the years. When I search Google for a date range as part of the query (not as part of an advanced search, more on that below...) it shows me exactly what I typed in. So if I search for "shoes 2005-2009" I get pages for that exact date range first, with some single year results for 2005 and/or 2009... but without any 2006, 07, or 08 results.

If I search for "shoes 2006" I don't get any of those 2005-2009 results, and instead get results from only 2006, or those with 2006 as part of the listed range (e.g. 2006-2007, 2004-2006, etc...).

In other words, when searched like that it just treats the date/s as part of the query like any other number - meaning you'd have to include that date in the on-page optimization.

However, a searcher can search within a custom date range on Google by going to Search Tools --> and changing "Any time" to "Custom range", then selecting the range from the calender. This isn't going to show those pages that are "optimized" for the dates, but rather the ones that were discovered by Google on those dates.

If you give us a little more detail about what you are trying to accomplish, what types of products you have, etc.... we may be able to provide more assistance with regard to best practices surrounding date-based archives/campaigns/products Vs "evergreen" landing pages that are frequently updated.

Good luck!