Sanity Check: NoIndexing a Boatload of URLs
-
Hi,
I'm working with a Shopify site that has about 10x more URLs in Google's index than it really ought to. This equals thousands of urls bloating the index. Shopify makes it super easy to make endless new collections of products, where none of the new collections has any new content... just a new mix of products. Over time, this makes for a ton of duplicate content.
My response, aside from making other new/unique content, is to select some choice collections with KW/topic opportunities in organic and add unique content to those pages. At the same time, noindexing the other 90% of excess collections pages.
The thing is there's evidently no method that I could find of just uploading a list of urls to Shopify to tag noindex. And, it's too time consuming to do this one url at a time, so I wrote a little script to add a noindex tag (not nofollow) to pages that share various identical title tags, since many of them do. This saves some time, but I have to be careful to not inadvertently noindex a page I want to keep.
Here are my questions:
-
Is this what you would do? To me it seems a little crazy that I have to do this by title tag, although faster than one at a time.
-
Would you follow it up with a deindex request (one url at a time) with Google or just let Google figure it out over time?
-
Are there any potential negative side effects from noindexing 90% of what Google is already aware of?
-
Any additional ideas?
Thanks! Best... Mike
-
-
Hi Michael
The problem you have is the very low value content that exists on all of those pages and the complete impossibility of writing any unique Titles, Descriptions and content. There are just too many of them.
With a footwear client of mine I no indexed a huge slug of tags taking the page count down by about 25% - we saw an immediate 22% increase in organic traffic in the first month. (March 18th 2017 - April 17th 2017) the duplicates were all size and colour related. Since canonicalising (I'm English lol) more content and taking the site from 25,000 pages to around 15,000 the site is now 76% ahead of last year for organics. This is real measurable change.
Now the arguments:
Canonicalisation
How are you going to canonicalise 10,000+ pages ? unless you have some kind of magic bullet you are not going to be able to but lets look at the logic.
Say we have a page of Widgets (brand) and they come in 7 sizes. When the range is fully in stock all of the brand/size pages will be identical to the brand page, apart from the title & description. So it would make sense to canonicalise back to the brand. Even when sizes started to run out, all of the sizes will be on the brand page. So size is a subset of the brand page.
Similar but not the same for colour. If colour is a tag then every colour sorted page will be on the brand page. So really they are the same page - just a slimmer selection. Now I accept that the brand page will contain all colours as it did all sizes but the similarity is so great - 95 % of the content being the same apart from the colour, that it makes sense to call them the same.
So for me Canonicalisation would be the way to go but it's just not possible as there are too many of them.
Noindex
The upside of noindex is that it is generally easier to put the noindex tag on the page as there is no URL to tag. The downside is that the page is then not indexed in Google so you lose a little juice - I would argue by the way that the chances of being found in Google for a size page is extremely slim, less than 2% of visits came from size pages before we junked them and most of those were from a newsletter so reality is <1% not worth bothering about You could leave off the nofollow so that Google crawls through all of the links on the pages - the better option.
Considering your problem and having experience of a number of sites with the same problem Noindex is your solution.
I hope that helps
Kind Regards
Nigel - Carousel Projects.
-
Hi Chris & Nigel,
Thank you for the considered responses. Good points about canonicalizing. A part I find frustrating is that the shared title tag across dozens or hundreds of pages will be across many different products/groups of products. So, the title tag is not a solid way to group canonicals.
Since the url patterns vary, I don't see how I could group these by which dozens or hundreds canonicalize to which one page, let alone make the change in Shopify other than one page at a time. My understanding is that this title tag manipulation is the only handle Shopify gives for making these bulk changes.
Gah!
So, here are my follow up questions:
-
How big of a negative is this in it's as-is state and how much better will noindexing most of the 90% make it Google Organic-wise? I ask because even the BS title tag to noindex project is a huge time suck.
-
If more is ever revealed about how to more efficiently group and canonicalize in Shopify, would adding the canonical after noindexing capture that lost authority later or would the previous noindex have irretrievably lost that?
-
Given all that, would you continue as I am?
Thanks! Best... Mike
-
-
Hi Mike
I see this a lot with sites that have a ton of tag groups. One site I am working on has 50,000 pages in Google caused by tags appending themselves to every version of a URL, the site only has 400 products. Example
Site/size-4
Site/womens/size-4
Site/womens/boots/size-4
Site/womens/boots/ankle/size-4
Site/womens/clarks/boots/size-4Etc etc - If there are other tags like colour and features, this can cause a huge 3 dimensional matrix of additional pages that can slow down the crawl of the site - Google may not crawl all of the site as a result.
If it's possible to canonicalse then that is the best option as juice and follows are retained - very often it would be the page with the tag lopped off that the tag should cite.
In extreme circumstances I would consider noindexing the pages as they offer very skinny content and rubbish Meta because it's impossible to handle them individually. I have seen significant improvement in organics as a result.
Personally I don't think it's enough to simply leave Google to figure it out although I have seen some sites with very high DA get away with it.
To be honest I am pretty shocked that Shopify doesn't have a feature to cope with this
Regards
Nigel
Carousel Projects.
-
Hello Michael Johnson and Mozzers,
I have seen Shopify do this a few times, though I do not have clients on that particular platform at the moment. It is frustrating. You're right to want to resolve this issue. Between duplicate content, authority conflicts, and an inflated crawl budget, one issue or another is bound to hold back site performance.
Is this what you would do? Not immediately, no. I want to see those pages canonicalized. That way, your preferred pages get all the juice back from their respective canonical link. Is this an option for you?
**Deindex request... and s_ide effects?**_ Canonical tags would make these part irrelevant (yay less work!). To be thorough though: I'd let Google figure it out unless you have strong evidence your crawl budget is maxed. And I don't see any negative side effects from noindexing duplicate content. If worse comes to worse, you have a good plan.
Shape that content,
CopyChrisSEO and the Vizergy Team
Got a burning SEO question?
Subscribe to Moz Pro to gain full access to Q&A, answer questions, and ask your own.
Browse Questions
Explore more categories
-
Moz Tools
Chat with the community about the Moz tools.
-
SEO Tactics
Discuss the SEO process with fellow marketers
-
Community
Discuss industry events, jobs, and news!
-
Digital Marketing
Chat about tactics outside of SEO
-
Research & Trends
Dive into research and trends in the search industry.
-
Support
Connect on product support and feature requests.
Related Questions
-
Product URL Optimisation
Hi guys, We are currently trying to add new products to our site but we are in a quandary on what type of URL structure to pursue. For example:
Intermediate & Advanced SEO | | michel_8
Product Name: Aspect Exfoliating Cleanser 240ml https://www.example.com.au/aspect-exfoliating-cleanser-240ml (including the size)
VS
https://www.example.com.au/aspect-exfoliating-cleanser 1.) Which is a better URL structure based on SEO 2018 and why?
2.) Is there any merit in removing the size from the URL key with the aim of attracting more traffic? Keen to hear from you guys! Cheers,0 -
How To Shorten Long URLS
Hi I want to shorten some URLs, if possible, that Moz is reporting as too long. They are all the same page but different categories - the page advertises jobs but the client requires various links to types of jobs on the menu. So the menu will have: Job type 1
Intermediate & Advanced SEO | | Ann64
Job type 2
Job Type 3 I'm getting the links by going to the page, clicking a dropdown to filter the Job type, then copying the resulting URL from the address bar. Bu these are really long & cumbersome. I presume if I used a URL shortener, this would count as redirects and alsonot be good for SEO. Any thoughts? Thanks
Ann0 -
Confused: Url Restructure
Hello, We're giving our website a bit of a spring clean in terms of SEO. The site is doing ok, but after the time invested in SEO, content and last year's migration of multiple sites into one, we're not seeing the increase in traffic we had hoped. Our current urls look something like this: /a-cake-company/cup-cakes/strawberry We have the company name as the first level as we with the migration we migrated many companies into one site. What we're considering is testing some pages with a structure like this: /cup-cakes/cup-cake-company-strawberry So we'll lose a level and we'll focus more on the category of the product rather than the brand. What's your thoughts on this? We weren't going to do a mass change yet, just a test, but is this something we should be focusing on? In terms of organisation our current url structure is perfect, but what about from an SEO point of view? In terms of keywords customers are looking for both options. Thanks!
Intermediate & Advanced SEO | | HB170 -
Attack of the dummy urls -- what to do?
It occurs to me that a malicious program could set up thousands of links to dummy pages on a website: www.mysite.com/dynamicpage/dummy123 www.mysite.com/dynamicpage/dummy456 etc.. How is this normally handled? Does a developer have to look at all the parameters to see if they are valid and if not, automatically create a 301 redirect or 404 not found? This requires a table lookup of acceptable url parameters for all new visitors. I was thinking that bad url names would be rare so it would be ok to just stop the program with a message, until I realized someone could intentionally set up links to non existent pages on a site.
Intermediate & Advanced SEO | | friendoffood1 -
Does having a ? on the end of your URL affect your SEO?
I have some redirects that were done with at "?" at the end of the URL to include google coding (i.e. you click on an adwords link and the google coding follows the redirected link). When there is not coding to follow the link just appears as "filename.html?". Will that affect us negatively SEO-wise? Thank you.
Intermediate & Advanced SEO | | RoxBrock1 -
URL language on Global Sites
Has anyone looked into a page not ranking as well because the URL is in English when the subdomain is geared for a different country and different language? I can defiantly see this taking away from the user experience, but didn't know if there was any concrete evidence or case studies that would show if it is a big deal or not for rankability? I know this is a backwards question to begin with because the priority over rankability is always UX, but there may not be a way to fix it unless I can prove it is a big deal.
Intermediate & Advanced SEO | | Ryan_Henry0 -
Does Google index url with hashtags?
We are setting up some Jquery tabs in a page that will produce the same url with hashtags. For example: index.php#aboutus, index.php#ourguarantee, etc. We don't want that content to be crawled as we'd like to prevent duplicate content. Does Google normally crawl such urls or does it just ignore them? Thanks in advance.
Intermediate & Advanced SEO | | seoppc20120 -
Could this URL issue be affecting our rankings?
Hi everyone, I have been building links to a site for a while now and we're struggling to get page 1 results for their desired keywords. We're wondering if a web development / URL structure issue could be to blame in what's holding it back. The way the site's been built means that there's a 'false' 1st-level in the URL structure. We're building deeplinks to the following page: www.example.com/blue-widgets/blue-widget-overview However, if you chop off the 2nd-level, you're not given a category page, it's a 404: www.example.com/blue-widgets/ - [Brings up a 404] I'm assuming the web developer built the site and URL structure this way just for the purposes of getting additional keywords in the URL. What's worse is that there is very little consistency across other products/services. Other pages/URLs include: www.example.com/green-widgets/widgets-in-green www.example.com/red-widgets/red-widget-intro-page www.example.com/yellow-widgets/yellow-widgets I'm wondering if Google is aware of these 'false' pages* and if so, if we should advise the client to change the URLs and therefore the URL structure of the website. This is bearing in mind that these pages haven't been linked to (because they don't exist) and therefore aren't being indexed by Google. I'm just wondering if Google can determine good/bad URL etiquette based on other parts of the URL, i.e. the fact that that middle bit doesn't exist. As a matter of fact, my colleague Steve asked this question on a blog post that Dr. Pete had written. Here's a link to Steve's comment - there are 2 replies below, one of which argues that this has no implication whatsoever. However, 5 months on, it's still an issue for us so it has me wondering... Many thanks!
Intermediate & Advanced SEO | | Gmorgan0