Welcome to the Q&A Forum

It's certainly beneficial to link to wherever is going to help the user navigate. However, I would be hesitant to put actual links to higher branches in this case because it's likely that the visitor path has already been there and for them to navigate away from the product page would only deter them further to completing your checkout process.

If your current site architecture (read: nav bar in this case) doesn't already link out to the category/subcategory page, then yes, it may help slightly with internal linking "juice", but it will likely be negligible, and you could/should compensate this with things like XML sitemaps anyway.

Another thing to consider would be the opportunity to cross/upsell to other products. It's fairly easy - depending on the CMS you're using - to build in a widget or some kind of functionality to help cross promote products on product pages. I mention this because it's a great alternative to giving people the option to click away from something they probably know they already want, as well as allowing you to increase your price per order by making it easier for them to add "related products" to their shopping cart.

Not knowing your exact situation, I would advise making the internal linking structure whatever is best for the path of the visitor. Make up any internal linking benefit that you think you might be missing by making sure you're XML sitemaps are in order, as well as interlinking wherever necessary in other parts of your site.

Hope that helps!

Related resources:
https://moz.com/blog/results-of-google-experimentation-only-the-first-anchor-text-counts
http://www.screenpages.com/about/articles/ecommerce-navigation

My suggestion would be to reevaluate the strategy of having that amount of pages in the first place. If you're dealing with a single type of software you're more than likely going to be working against yourself, not only in terms of cannibalization, but user experience as well.

Instead, beef up the content on a single page (ie, www.domain.com/product) and make sure to include as much coherent information, as well as the variations, that you can. The engines look for these variations anyway, so as long as you're spoon feeding the information to them on this page, they should be able to pick up that it's the same product and rank it accordingly once there are other signals in place (links, anchor text, etc).

Having said that, if you do have significant enough differences in the software to warrant multiple pages, then you simply need to come up with better descriptors/keywords, and build out the their specific. Make sure that the content is unique enough, but not spilling over into what the other, similar pieces of software do. Otherwise you might still run into your cannibalization issue. Ranking these in the beginning may be tough, but as Google recrawls these pages over and over they will start to pick up more differences and "understand" them better. I'll also echo my earlier suggestion - make sure the necessary signals are in place here too. You might be able to help Google understand the differences in these pages quicker by being a little deliberate as far as anchor text is concerned.

Hope that helps!

I'll echo EGOL's response, just because a link is coming from a specific place doesn't make it inherently bad. Relevance and context is certainly more in line with what you want to look for when judging value of a link.

Hope that helps!

Hi Harry,

Logan is right, this happens a lot when new projects get started, especially when a site is newer.

The easiest way that I've found to combat this (in addition to what you've already done) is to build some almost "over optimized" links to each of the pages in question. When our team does this, we make sure 1) to include exact match anchor text of links that are built, 2) the links and linking domain itself are extremely relevant to the destination page, and obviously, 3) they are pointed at the very specific pages in question that you're trying get ranked properly.

I certainly wouldn't be this blatant all the time, but when trying to "separate" these pages in the search results I would definitely make sure this is on your checklist.

The search engines are incredibly "intelligent" but they're still machines and can't necessarily infer what page needs to be ranked without the proper signals. So making sure the on and off page signals are there to provide as much context as possible is really important.

Hope that helps Harry!

Hi Vanessa,

First off, really awesome question, I definitely had to give this one some thought! I would do both...

You're right about there being pros and cons, but if it were me I would create a dedicated landing page for each contest then 301 them to the main /contest page after it was over.

The reason for my preference is because you may be more likely to earn certain links from certain places if you have a very specific landing URL vs. a generic one. I know you mentioned changing the "catch all" page each time, but you lose the ability to change the URL in that scenario. If it's always going to remain /contest, this might be a turnoff (albeit very slight) to someone that might have otherwise linked to you if it was /unique-content-title instead. Not to mention, if there are any semi-shady domains that end up linking to this unique page, the URL specifically is another signal that might reinforce the idea of "Don't-penalize-me-because-their-spammy-site-is-linking-to-me-I'm-just-running-a-contest" to the engines because of its onpage relevance (this also assumes it's a relevant spammy site linking to you in the first place).

All that being said, you would lose about 10% of link equity anyway by 301ing the unique link every time, and the scenario of getting one bad link that ends up hurting you is unlikely, so it really could go either way. This decision might also be swayed by how hard it is to get something redirected in the first place. If you're in a "typical" corporate environment, it may be an act of congress to get a 301 pushed through instead of just changing the body content of an existing page.

Hopefully that helped Vanessa and didn't just generate more questions...

It sounds like first you would need to know how the other states are using their sites (folders). This will help you answer the canonical vs. unique content question.

For example, if the individual states are going put their own marketing efforts behind their piece of the site, you definitely don't want to canonical their section as power would be transferred away. However, if they're not actively marketing their section (building links, blogging, whatever), AND your client doesn't mind the individual states not having a lot of ranking power, then I'd say using canonicals within each section is a viable option.

Let's assume that for whatever reason you can't use canonicals. Then it seems like you can either 1) suggest some type of plan for the states to start building out their own content in their areas so content is more unique all the way around, or 2) have the association (or you) do it themselves. Regardless, there's elbow grease involved.

That's partially a question of how soon the linking pages get indexed/recrawled, but in general it's very fast. On our last project where we had an issue like this it only took ~4 days before our pages were showing up correctly in the search results. We were targeting two pages specifically, and built around 8 links per page.

Keep in mind though that the homepage is almost certainly going to be stronger than any of your other pages, so rankings might slip a bit at first when the engines start to attribute the pages correctly.

Hi Lise,

Generally this isn't something you should do unless you just made the page and you want to change it (as in the case of a misspelling in the URL, for example). Like Andy mentioned, you're essentially just creating another page from scratch.

I'd also suggest creating a 301 redirect from the old page to the new one. This way any link equity remains intact.

Here's a good Moz post on this topic: https://moz.com/blog/should-i-change-my-urls-for-seo

In my experience it's rarely the best solution to break up a site (or start out with separate sites for that matter). The cases that generally warrant having multiple properties don't usually have anything to do with the site itself, but the business. An example would be a medium fast food chain that begins allowing franchise operations, but wants each location to now have it's own website.

A big reason for this so rarely being a good option is because you're splitting assets, plain and simple. Domain authority, link equity, brand awareness, inevitable loss in rank for certain pages, not to mention the technical work on the back end that would have to be done (site architecture rearrangement on the new sites, 301s, etc). And this isn't just in the beginning either - you're going to always have to do things in 3s once you make the change. 3 link acquisition campaigns, 3 content campaigns, 3 social campaigns (not to mention the individual accounts), and I can't even begin to think about your PPC, whew! That's a ton of work.

In my opinion, you're much better off reorganizing your current site for long term success.

References:
Splitting a Site into Two Sites for SEO Purposes TCS Case Study: Why We Combined Two Sites Into One Web Content Management Strategy

Hi,

I wouldn't worry too much about this. It's not ideal to have so many duplicates (especially if they're on the same page), which is probably why this is getting tripped in the audit tools, but it's not something I would lose sleep about. At the end of the day, headings are for structuring content for the user, so if it makes sense, do it.

It's mostly about multiple headers on the same page, but here's Matt Cutts describing what's good/bad about this practice: https://www.youtube.com/watch?v=GIn5qJKU8VM

Hope that helps!

Nailed it.

Hi,

Part of this depends on how the backend of the website is structured. If it's easy enough you could make mass changes to the main celebrity page and have those changes take effect on your other pages as well.

Obviously this doesn't change the fact that you would still have 20K page to update. With that in mind, another scenario might be to put canonical tags on the news, videos, and image pages that point back to the main /celebrityname page. This depends on how you want your pages to show up in search results, but that would help cut down on the duplicate content, as well as putting ranking "power" in one main place without having to make a ton of manual content changes.

Hope that helps!