Learning SEO by Looking at Anomalies

Sharing is caring!

Sometimes one of the best ways to learn is to question why you see something that you possibly shouldn’t be seeing. This can produce more interesting lessons than even digging into things like patents and whitepapers. For instance, I published a post last night, October 4th, 2016, on Context Vectors. On Twitter this morning, Dan Shure had a question about the snippet for my post, wondering why the snippet date was April 10, 2016. I knew the answer immediately.

Dan’s Questioning Tweet:


When I researched the post, and found the patent I wanted to write about, I used the WIPO Patentscope website to find the patent. I copied information from the WIPO pages, including the formatting of the Dates the patent was published and granted:

Dates from the WIPO site

I copied those dates into my post in exactly the same formatting found on the WIPO site DD/MM/YY instead of the way dates are formatted in the US MM/DD/YY:

Dates were used in my post

It appears that Google decided that it should take a lesson from the context of my blog post, and read the date of my post using DD/MM/YY instead of MM/DD/YY:

Actual Post Date

Which answers the question that Dan asked about why Google decided that the date in the snippet should be April 10, 2016, instead of the date that it was published on October 4, 2016. Google may have paid more attention to the context of date formatting than it possibly should have in that post. It does show that Google is paying attention to context, though.

There were some others involved in a discussion of this anomaly this morning, and I wanted to give a shout-out to them for getting involved (thanks!):

Aymen Loukil
Steven Weldler
Tylor Hermanson

I’m not sure if it will help much, but I have changed the Formatting of the patent dates in the post about context vectors to the Americanized format of MM/DD/YY. However, Google is still showing a snippet date of April 10, 2016.

Aymen Loukil suggested I try using DateTime Schema. It looks like that is to provide information regarding sites that provide schedules, and Schema.org tells us that it is used on “Between 10 and 100 domains”. Nothing on that page seemed to explain how to explain the formatting used by Dates.

Funny to learn about how Google might use context to format a snippet date based on how the snippet is about may format dates. I decided to write about Dan’s question and its answer to keep it in mind and learn from it (I suspect we won’t see a patent just about how Google decides how to format dates in snippets.

I’ve written some other posts about patents from the WIPO pages and used the European formatting dates from those. So I should go back and look at the snippet dates Google may have given those…

Sharing is caring!

48 thoughts on “Learning SEO by Looking at Anomalies”

  1. Hey Bill, FWIW I declare my blog posts to be of type schema.org/Article and then provide datePublished and dateModified in ISO 8601 date format (as per schema.org/Date). Google’s seems to understand this pretty well (though I’m not sure if it favors this structured data over in-post dates).

  2. Hi Aaron,

    Thanks. I’m not sure I even thought about whether or not the formatting of the dates I included in my post from yesterday would influence the formatting of the snippet date that Google showed for my blog post. I have now learned that Google does pay attention to information like that, which is why I am sharing the information.

    We can see that Google may pay attention to the dates that are shown on a post, but might ignore the country of origin of the post when deciding upon the date format to use to interpret when those in-post dates actually refer to, if the post refers to dates that are formatted differently. I probably need to find other posts I’ve written that have included WIPO patent information, and see if the snippet dates showing for those also uses european formatting in snippets for those posts.

    If I had a site that published schedules for transit, I would probably want to find a way to format dates so that I was certain that Google understood the formatting I was using. Snippet dates, though? I wrote and published the post yesterday, which wasn’t April 10, 2016; I didn’t expect Google to change the date format that it had been using for most other posts on my site because I didn’t change the formatting of the dates I received from WIPO. Now I understand the risk of not doing so. ๐Ÿ™

    Maybe including the Structured Data information in addition to the in-post dates may be a good idea.

  3. Hi Aaron,

    Thinking about this again, I wonder if I had the right structured data, and included European formatted dates from WIPO in my post if Google would choose the formatting it sees in the post over the information it sees in structured data. I don’t know which it would choose. California is a long way from Europe, so I don’t know why they decided to use european formatting.

  4. Question: What exactly does this say about Google’s ability to track content within context? Would you say it proves that contextual backlinks are given more weight then say a random link posted without any content, or a link thrown in a side bar?

  5. Hi Cliff,

    Thank you for sharing your questions. This patent describes a process by which a search engine can better index content found on the Web and assign contextual meanings to terms or phrases or words based upon finding those in a knowledge base such as an online encyclopedia, and assign a weight to the term(s) based upon how often it shows up in such sources. The combination of the contextual meanings (from different subject matter domains) and weights applied to those meanings is referred to as a “Context Vector.” The patent isn’t about links or backlinks, and doesn’t refer to contextual backlinks at all, so comparing it to a random link or a link in a sidebar isn’t a consideration here at all. The idea behind this patent is to better index content found on the web, and to apply contexts to queries when searches are performed to make it more likely that a searcher finds what they are looking for.

  6. Out of curiosity, does a post with earlier dates typically rank higher? For example, 2012 vs a 2016 post, will the 2016 post rank higher being that its newer information, or is that not factored in with seo?

  7. Hi Bill, Thanks for this useful information, this will help a lot of beginners like me.

    Thanks Again.

  8. Hi Rasmita,

    You’re welcome. If you do see something odd in Google’s search results. and you can capture a screen shot of it, do so. And then ask about it, in places like Twitter. It is a good way to learn. Sometimes you can catch Google experimenting on something, and it’s good to see that happening, too.

  9. Interesting to see how Google really drills down to content/context and not just url/title/description/H1. Really need to be on your toes while writing anything these days!

  10. Hi Andrew,

    It is interesting seeing how Google can get a sense of the meanings of words on pages, and the context of those words. If what you write is about something, in a particular context, it should be easier for Google to find it. I’m liking the idea of context vectors.

  11. Wow, that is just fascinating to read. For all the simplicity of the Google search bar, there is so much engineering going on behind the scenes. I wonder how complicated the algorithm is if they pay so much attention to a date within the content!

  12. hi bill,
    really thanks for sharing this valuable information
    this will help lot of begginers like me…
    thanks for sharing. and keep us updating like this

  13. wow great post friend
    i really impressed by your article
    what a explanation
    thanks for sharing such great article keep us updating like this
    thank you so much friend

  14. Hi Bill,
    Thanks for sharing your informative and amazing post, Iรขโ‚ฌโ„ขm benefit from sufficient. I hope everyone will be benefited if they read all of your post. Please write more for us as we learn more. Thanks again,
    Take Care

  15. I like your post!!!!!!
    it’ very helpful for my business
    if you have some more knowledge about it, please share

  16. A very good article. This will help a lot of beginners like me.
    Thanks for sharing.

  17. Thanks Bill,
    i saw Aymen Loukil speaking at a SEO meeting last month in Lyon (Fr), he’s a good expert about schema.org. anyway thank you for sharing this with us.

  18. Hi Franklin,

    I have been hearing positive things said about Aymen Loukil. I have seen him on Twitter, and I am going to have to hunt down things he has written about Schema. Thanks!

  19. Hi Bill, quick question – do you think implementing a date time schema would help in ranking newer content in any way or is it just for context purposes?

  20. Hi Alnoor,

    I don’t think that a date/time schema would solve the problem I have with the format of the date the post was published being read wrong by Google. It probably doesn’t have any impact on the rankings of that page either.

  21. Hi Bill, Thanks for this useful information, this will help a lot of beginners like me.

    Thanks Again.

  22. Does that mean if I publish a post in 2013 and update it in 2016, will that post show up higher in search results?
    Great post once again. Thanks Bill

  23. Hi Flypa,

    No, changing the dates on post that you made shouldn’t have any effect on the rankings of those posts. However, if Google notices that a post you wrote in the past starts getting a lot of new links pointed to it, it might take those new links as a sign that people are interested in the content of your post, and might consider it timely or evergreen.

  24. Thanks Bill for helpful content, I remember having difficulty with schema markup for blogs and articles had few errors with regards to dates. Keep up producing helpful content for the community. Best regards

  25. This website is very informative to read. I am a huge follower of the things you talk about. I also love reading the comments, but it seems like a great deal of readers need to stay on topic to try and add new things in the original topic. It and also added your RSS feeds, so when I have time I will be back to read a great deal more.

  26. Hi Roy,

    It’s more enjoyable to me when comments are on topic, and add more to the original topic. I hope to see you again, and will look forward to your comments. Thanks. ๐Ÿ™‚

  27. Great article. I love how you discovered this gem by paying attention to anomalies. Reminds me of the confirmation bias — people often look for things that confirm their views rather than looking for disconfirming evidence, even though a disconfirming instance may provide you with much more information than you otherwise would have had (including proving that your initial thoughts may have been wrong).

    I’ve bookmarked your page and hope to read more good articles from you in the future!

  28. Hi Marco Z,

    I had no views to confirm when I was writing this post – just the thought that if I keep my eyes open for something unexpected, I might see something unusual and learn something about how Google behaves sometimes. ๐Ÿ™‚

  29. Bill, I must say that you have done a fab job by sharing something unique and interesting topic, this gonna be a really helpful to the startups and beginners. I really enjoyed the post and going to share this in my circle.

  30. Great insight Bill. I have an article request: Could you help as explain how to trigger knowledge graph result in Google? I struggle with this greatly.

  31. Hi Emil,

    Knowledge graph results are triggered by whether or not you have an entity in your query that Google recognizes, and has a knowledge panel result for. Sometimes, in addition to the entity in your query, you may also need to include an additional term to cause a knowledge panel to show up, which I’ve seen Google refer to as a Context term. I will likely be writing about that in the very near future.

  32. Hey Bill,

    Great info on the snippets. I’ve been trying to reverse engineer some top local listings for clients who want to rank above the organics, and the snippets can sometimes be unwieldy. I like your process of reverse-engineering though to figure out what the program will pick up from each part of the page. Thanks again for sharing and I’ll keep my eye out for anything else you post on snippets.

  33. Hi Mitch,

    This post was really just a lool at one aspect of snippets, that Google might pick up on formatting used in dates, and have that influence how it handles dates such as the posted upon date. I did write a new post recently that includes a wired article that tells us Google is using a new sentence compression algorithm to generate answers in featured snippets. Those may in some instances rewite content on people’s pages that may answer questions in queries – I guess we may need to keep on eye on featured snippets, and see if it does rewrite those now. the newer post is this one:

    Answering Featured Snippets Timely, Using Sentence Compression on News

  34. The end goal should be to provide those searching online with exactly what they requested…It sounds like context vector will help in that regard by doing a better job of ensuring Google returns what consumers are expecting to see. No claim to fully understanding it but I do like it when my searches net what I am seeking.

  35. Hi Anita,

    The Context Vector approach is interesting, and should provide smarter search results if it can better understand the meanings of words being searched for and used on web pages. I’ve been trying to create content that takes better advantage of contexts.

  36. Hi Bill,

    Wow! Good to know google’s paying attention (a wee bit scary but good). We in India use DD/MM/YYYY and i think the only time we change it is working with a foreign client. This has got me peering a little harder at my search result.
    Learned something new today!


Comments are closed.