Keep It Fresh

Keep It Fresh
Don't let your content wither. Monitor old content on a weekly basis and update to keep it fresh. The results work for search and for your site visitors.

#contentmarketing  

Originally shared by David Kutcher

Is the searcher seeking fresh content, and how fresh is your content?

That's at the heart of this patent by Google, discovered and dissected by Bill Slawski.

"A patent granted this week for Google discusses the concept of “Freshness Based Ranking” and is named that. It not only looks at whether the query in question might prefer a fresher result, but also older results, and could boost both types of results based upon what seems most called for."

This fits back into my Maintenance Monday activities, discussed ad nauseum I'm sure, that your site and evergreen content can benefit by graceful aging, but also through continued refreshment with new contributions, current links, etc.

Keep that original publish date, but also proudly display that "last updated" date to your user and to the search engine (via schema).

Read Bill's post on the patent, or my own (updated!) post on content life cycles at http://www.confluentforms.com/2013/09/blog-content-life-cycle-dividends.html

H/T: Mark Traphagen
http://gofishdigital.com/new-google-freshness-based-ranking-patent/

Comments

  1. David Kutcher, it would be my understanding that new comments would contribute to the link freshness? Are you saying that the last update timestamp is what they are looking at?

    ReplyDelete
  2. Denver Prophit Jr. I'm not sure exactly what they're using to make that freshness determination, but lastUpdated is a valid part of schema for a reason IMO. So adding content to a page, or making revisions that generate a new lastUpdated, and that Google can see by comparing new to indexed versions, would likely result in a reindexing and freshness score.

    ReplyDelete
  3. The core use of a sitemap for G saves time and bytes, two things G likes.

    ReplyDelete
  4. Padraig Ó Raghaill in this case the sitemap does partial duty, but only partial.

    ReplyDelete
  5. Why only partial I thought most XML generated maps have an updated and a last updated string? I was also under the impression G had a freshness rank already in place and was/is used against news items and content that is expected to have a higher relevancy if fresh. The patent looks to be expanding on that David Kutcher

    ReplyDelete
  6. Padraig Ó Raghaill​ for one thing, if google trusted your date they'd be very gullible. Instead they most likely use it to compare vs the already indexed version to see if it is indeed fresh, and how much it was freshened.

    Additionally, it is important to let your readers know it is fresh, and few readers will be examining your sitemap to check freshness, instead looking in page for that information.

    ReplyDelete
  7. That is not what was referenced David Kutcher 1 Google would and does trust the map and even Matt Cutts gave insight into that some years back. Two dates on  evergreen is a double edged sword and to top it off G still looks at both copies and is irelavant to was was also just highlighted. So in essence, as I said looks to be a further resource on their existing freshness rank. Plus no one (by the way) readers needs any fancy Schema or similar just a valid and and well structured site map following Googles recommendations . If you can save Google the visit they will thank you for thinking about their bytes. Or as Matt Cutts put it, Google likes to be lazy and if you make her sweat without good need she may just not thank you for it.

    ReplyDelete
  8. David Kutcher, well the vocabulary definition isn't meant for indexing. Don't you think it's meant for the knowledge graph?

    *Packet Headers*
    https://en.wikipedia.org/wiki/List_of_HTTP_header_fields.

    Software should update the last-modified timestamp anytime a new comment is added to the page. That triggers browser caching to revalidate.

    The file size would also be a signal for browsers to revalidate cache.

    That's why I began to wonder about comments on the page creating freshness.

    ReplyDelete
  9. Padraig Ó Raghaill I would encourage you to read the search quality ratings in regards to dates and handling of freshness. http://googlewebmastercentral.blogspot.com/2015/11/updating-our-search-quality-rating.html

    ReplyDelete
  10. Denver Prophit Jr. I think it would then depend on the commenting system you use and how that system renders to Google, perhaps using Google's tool in WMT to test. But yes, perhaps, as I know g+ comments in my blog posts do add to the on page content. However, I'm also convinced that on page content is broken into zones by the algorithm when evaluating the level of modifications that generate freshness.

    ReplyDelete
  11. I added a WP ticket https://core.trac.wordpress.org/ticket/34775#ticket to update the SQL record when a comment is accepted.

    As far as google results filtered by xxx timespan, does it rely on date created or date last updated when rendering snippet results for that type of filter and freshness?

    The patent summary gave an example of a video from a band being created and emphasis on timeliness of the query results.

    Is there a hint that last update might receive less weight than date created and specific to the query intent?

    ReplyDelete
  12. Denver Prophit Jr. I'm not sure, and will need to research further. Bill Slawski, thoughts? I don't think that an accepted comment should change the last updated date of the originating post, as that post was not updated. That being said, each comment should have a published date on it, with comment schema being used. Again, Google is comparing the rendered page against the indexed page.

    ReplyDelete
  13. The ticket will get clarification. I suspect the file length is different and the browser cache revalidates. The last-modified header response should reflect the latest change to the entire page.

    ReplyDelete
  14. David Kutcher, again comment schema is only going to affect machine learning. Not the RFC compliant documentation - http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html

    ReplyDelete
  15. I think maybe both can be right. Bill should clarify his study to that aspect.

    ReplyDelete
  16. Nothing new there to be fair David Kutcher my understanding is sound.

    ReplyDelete
  17. Just checked out the Moz article it flows into exactly what I was citing with the new patent extending, updating, contributing however you want to look at it, extending how they have since 2003 referenced in the article, looked at query and freshness.

    ReplyDelete

Post a Comment

Popular Posts