Want to slowly kill your content on Google? Simply use a directory structure with dates

Juan Gonzalez
.I studied Regional Studies of Latin American at the University of Cologne - Germany, majoring in "Business Informatics“. I also studied Business Administration and currently I’m doing a Master in International Business Administration. I feel a fascination with SEO and the people who make it possible.
Artikel teilen
15. September 2016 17 Comments

Visibility Index for Techcrunch.com on Google.com

The main reason to avoid dates within your directory structure is explained on page number 8 of the Google’s Search Engine Optimization Starter Guide:
Simple-to-understand URLs will convey content information easily Creating descriptive categories and filenames for the documents on your website can not only help you keep your site better organized, but it could also lead to better crawling of your documents by search engines. Also, it can create easier, "friendlier" URLs for those that want to link to your content. Visitors may be intimidated by extremely long and cryptic URLs that contain few recognizable words.
If the dates constitute a really relevant piece of information for the user, I would keep them. For all other cases I would advice against them, as in doing so, you are likely going to kill your content on Google.

What would work better for search engines and their users? /2013/dec/05/ or just /Nelson_Mandela

Visibility Index for Huffintongpost.co.uk on Google.co.uk

The Huffingtonpost.co.uk has created much more content about Nelson Mandela than the Wikipedia. For example, I was able to find 29 individual news articles for the Huffingtonpost.co.uk that use Nelson Mandela within the URL. Interestingly enough none of them rank on the first page of Google’s search results.

For Theguardian.com I even found 443 individual news items that use Nelson Mandela in their URL. Here, only 27 URLs rank in the top 10 and all of them are long tail keywords, which usually have a very low traffic volume. The content on Theguardian.com might very well be more reliable and verified than that on the Wikipedia, but that will not help you if you cannot find it on Google because every piece of content on /world/2014/ and /world/2015/ does not rank well anymore:

Visibility Index on Google.co.uk for the directories theguardian.com/world/2015/ and theguardian.com/world/2014/

If you search for Nelson Mandela on Google.co.uk we quickly notice that nine of the results use user-friendly URLs. You have the Wikipedia at position number 2, a piece of archived content by the BBC at position number 4 and Theguardian.com at position Number 10 (actually 11), within the “related news box” – which is not the same as Google’s “In the news” vertical:

First search result page on Google.co.uk for the keyword “Nelson Mandela”

Having the Date in the URL is not the cause for the losses but a symptom of a weak information architecture. Both the Wikipedia and the BBC articles rank better than TheGuardian, not only because the URL suggest what the content is about but because the content is useful and timeless.

Don’t benchmark within your industry

One of the patterns of the news and information industry is that they all tend to do things similarily. Even the way they handle their information is almost identical, just like a printed journal. All the resources invested in research and editorial work disappear on Google after one year. Content that you have already paid for must not necessary be lost on Google after such a short time.

It is natural that fast lived news tend to fade but the same could be said of fashion trends – and online shops can deal with those very well. The number of topics that a journal will use, like Nelson Mandela, James Cameron, Hillary Clinton, Brexit, etc., are quantifiable because they are almost always the same, just like general product categories for e-commerce sites.

Wikipedia has 5 million articles for their English version (not topics!) and can deal with the above problem well. When we look at e-commerce sites it gets even crazier. Amazon, for example, has to deal with 60 million different accessories just within their Electronics and Photo category. This should give you a good idea that, if you really want to, you can indeed deal with a very large amount of topics/products.

The Search Result number 10

I asked John Muller in a German Hangout on February 25, 2016, why the search result number 10 is composed by 3 different snippets. He told us, it is just a sort of “Related News Box“.

The Theguardian’s search result for the keyword “Nelson Mandela“ is their obituary. We really are talking about a very good, high quality obituary here, which, in my opinion, is much better than the corresponding part on the Wikipedia. I have another very good example for amazing content: Just search for “greenland“ on Google US and you will probably find this:

Search Result number 10 for the Keyword “greenland” on Google.com

Journals are able to create great content. Just click on the search results for the Nytimes.com and Rollingstone.com. Next let us take a look at the history of the keyword “melting glaciers” for the Nytimes.com:

Keyword history of the keyword “melting glaciers” for the Nytimes.com

Compared to the general downward Visibility trend for news items, we can see that this article has had its ups and downs but now ranks within the Top10 – actually for a total of 36 keywords related to “greenland” and “melting glaciers”. All 3 snippets for this keyword, within the “related news box”, are very good alternatives to the Wikipedia. It is not healthy for information seekers to get the Wikipedia as the only source of information.

Lastly, would you have ever imagined that Rollingstone.com would compete with the Nytimes.com and Nature.com on a topic such as melting glaciers? This Rollingstone’s article has a simple-to-understand URL, as well as timeless and and great content.

I hope you like it!

15. September 2016, 21:06

so is this really due to having dates in the URL or is it due to the fact that most pages with dates are news and most pages without are static content. I would argue that old news, or his obituary are less relevant for a search on his name than a profile page or a summary page about him. I think this has more to do with the actual content of the pages, is it old news story or is it static (very different) than it does with the URL structure.

Juan Gonzalez
7. October 2016, 15:54

Hi Ryan,

as we said, “Having the Date in the URL is not the cause for the losses but a symptom of a weak information architecture”.

Not all news have to be old news and not all static content has to be fresh. As Google would say: “It depends”.

Best wishes,
Juan

20. September 2016, 19:12

Interesting findings Jaun 🙂

Juan Gonzalez
7. October 2016, 15:55

Hi Andy,

Thank you very much! I’m glad you liked it 🙂

Kasia Lorenc
20. September 2016, 20:31

Agreed, dates in urls usually don’t make sense, for SEO or usability, but I’ve found that more and more sites don’t include dates on pages at all. Sure, there are many cases where they’re not necessary. But there are just as many situations where the publish date is a crucial piece of info, especially if the article is being referenced.

Take for example the Rolling Stones piece you linked to on Greenland — there’s no clue on when that went live. The author refers to “a few weeks ago,” “one week in July last year,” “earlier this year,” etc. It’s infuriating not being able to know if this is a brand new piece of content that just went live or if it was published last year, or another year altogether. Looking at the page it’s anyone’s guess.

According to the Internet Archive the url was crawled for the first time on Dec 2014! This means this story is 2 years old and it’s very likely that much as changed since it was written and published. Yet readers won’t know this.

I get the dismissal for dates on websites, for the reasons you listed above and more, but it would be nice if publications that are striving to offer valuable information would at least try to stick to some type of standard on this.

Interested to hear your thoughts.

Juan Gonzalez
7. October 2016, 16:01

Hi Kasia,

thank you for your thoughtful comment. I really like it 🙂

I do agree 100% with you that dates should be part of news articles and the like. The dates should be accessible on the page itself, but have no place within the information architecture of the website.

Best wishes,
Juan

20. September 2016, 23:23

Good point… Better go change my permalinks now…

Juan Gonzalez
7. October 2016, 16:13

Hi Kalon,

I’m glad you liked the article 🙂

Yeah! The users won’t search for “September 7 2016 iphone 7” but for “iphone 7”.

BTW: Don’t forget the 301 redirects https://www.sistrix.com/blog/seo-basics-never-relaunch-a-website-without-redirects/

Best wishes,
Juan

21. September 2016, 11:33

Hi Juan, nice one. In general I agree, those URLs with dates don’t look good. And the actual content gets buried too deep in a complicated site structure that doesn’t add any real value.

Then again, publishers like the Guardian have massive amounts of content, which needs to be structured. Since a large part of that content is time-sensitive, it makes sense to categorize by date (good for Google News as well, a potentially huge traffic driver).

Anyway, whatever you use to structure URL’s of individual articles, it’s vital to group content by topic, bringing together all the loose parts and create a great landing page. You can do this well with categories or tags. Looking at the Mandela example, I’d say that this page https://www.theguardian.com/world/nelsonmandela is (or should be) the most important NM page on the Guardian website. So why is Google not showing this in a UK search for the short tail keyword ‘Nelson Mandela’?

My take: the obituary page has a single, clear purpose. It will have gained a lot of incoming links and shares. In contrast, the tag page isn’t focused enough, there’s too much going on. The articles grouped are sorted by date, and a wide range of topics is covered. In many articles Mandela is only mentioned, they aren’t really ABOUT Mandela.

If The Guardian wants to change the tag page into a true ‘Nelson Mandela’ landing page, they should:
* write a proper introduction
* feature the obituary at the top
* train editors to be more selective in adding tags to articles.

Juan Gonzalez
7. October 2016, 16:20

Hi Arne,

Thank you for your detailed comment. You are spot on!

The category page about Nelson Mandela is surely something the Guardian can do better, as they are already doing tremendously good work.

In order to keep Google happy, it makes sense to invest more labour than might actually be necessary to create and especially curate category-pages. Even then, they will usually be those pages on a domain which pull in the most visitors with the least bit of effort.

Cheers,

Juan

22. September 2016, 16:33

Great Study Juan! It’s a shame that platforms like wordpress have this URL structure on by default. On the other hand, it would be curious if you could show graphs from sources other than major newspapers (like Mashable for example) and show growth graphics from websites that don’t use directory structures based on dates. Thank you!

Juan Gonzalez
7. October 2016, 16:22

Hi Daniel,

thank you for your comment!

We do have a number of examples and I will write about them in a few weeks.

Best wishes,

Juan

25. September 2016, 14:33

Interesting huh…

Have you ever tried to test what will happen when you remove dates from URL and POSTS at the same time (and from snippets too) ?

Juan Gonzalez
7. October 2016, 16:24

Hi Zgred,

thank you for your comment.

That is something we have not yet tested for. Please let me know if you tried it out.

Cheers,

Juan

26. September 2016, 20:29

To me, there are really two parts to this. If you are using dates in the URL already and have a great following or if you’re just starting out. If you’re just starting out using a simple /blog/title is the easiest way for Google to index and gives your /blog the most juice. At PowerObjects, we have been using the dates since the inception of our blog and continue to see traffic to our older posts. In our top ten organic landing pages, our posts are laid out like this from 10-1 by year: 2013, 2011, 2014, 2012, 2015, 2016, 2015, 2013, 2015, 2015. So while there may be a loss in overall file 2013, there are still posts being used in Google search going back to 2011 because of the content. Keeping these helps us strategize on what works during a certain time and what doesn’t. If it’s evergreen like those above we’ll continue to update or direct people to an updated post on the subject.

I’ll wrap this by saying nothing anyone has done to a single site is guaranteed to work for another. If your site is healthy and not losing traffic in a large amount don’t touch it. Find a way to utilize best practices with your current site layout to gain the most search impressions and create content that produces good numbers.

Juan Gonzalez
7. October 2016, 16:43

Hi Chris,

thank you for your comment!

You are right that older content may still be very useful for visitors today. Please note that in this article we are not talking about the content itself but the information architecture of a website.

The examples from the Guardian, the Huffington Post and TechCrunch all show us that there is still much room for improvement and probably for much more traffic.

Best wishes,

Juan

[…] Want to slowly kill your content on Google? Simply use a directory structure with dates by Juan Gonzalez (@JuanGonzalez_JP)  […]