Feel free to “like” this post

Recently, I’ve read a few articles published about the “Like” button on Facebook.  These have been done after some “research” conducted by the author using their own Facebook page as a source of data.  One, published by Wired, is done in such a way that the author decides to “like everything” for a few weeks and see how his news feed changes.  Another, published on Medium.com, is exactly the opposite: the author decides to not like anything for a few weeks and see how this affects things.

In both the articles, the authors ultimately make the case that the like button seems to be this magic button that drives everything in your news feed for better or worse and thus you must be very careful about if/how you use it.  It’s almost like they’re elevating the “like” button to some sort of medieval sorcery button.

Should be the

This should be the header image for these articles.  Or maybe replace “Here Be Dragons” with “Like At Your Own Risk.”

The problem with both of these is that they pick out one facet of an overall algorithm, try to do something to the extreme of it, and then report their anecdotal “findings.”  The way these articles are written have a level of encouragement to the readers to use the “like” button carefully.  And that’s fine: you probably shouldn’t go click “like” on everything because that’s simply a stupid thing to do.  I mean, does anybody really need to tell you that clicking “like” on things you don’t like is probably going to adjust your news feed toward things you don’t like?  That seems like an exercise in tautology.

The unfortunate thing is that, since these articles have come out, I’ve seen people saying things like “well I guess I won’t click ‘like’ anymore, I’ll just start using comments instead so I don’t fall victim to this ‘like’ button thing.”  Well guess what: comments are counted by Facebook too!  And that’s where these articles really lost me in a sense: they tested 1 variable but completely ignored other variables that Facebook is tracking.

Facebook's old EdgeRank algorithm

Facebook’s old EdgeRank algorithm

Facebook used to provide an approximation of their news feed algorithm which they used to call “EdgeRank.”  There were weights for different types of things (including, but not limited to likes).  Commenting, tagging, creating posts, etc… those all were part of this algorithm and they all had their own weights.  So yes, “liking” certainly would affect things, but in the Wired article, the author doesn’t mention anything about what he commented or tagged or created during the period.  So how much of what he saw was affected by the “liking” as opposed to other things he was doing?

How many licks to the center of a Tootsie Pop? Or how many likes to the center of the Facebook algorithm?

How many licks to the center of a Tootsie Pop? Or how many likes to the center of the Facebook algorithm?

Since the “EdgeRank” algorithm was replaced circa 2013 and replaced with an algorithm that takes into account “100,000 individual weights in the model that produces News Feed,” it’s only gotten more complicated.  They no longer provide real insights to what/how they’re tracking, but I’m sure that’d be pointless anyway as I’m sure it’s evolving still to this day.  So I’m not going to end this with pretending to know exactly what Facebook’s News Feed algorithm is or hint at ways to sidestep it or fool it into “giving you better results.”  That’s for the Facebook engineers to work out.  However, it is simply worth noting how naively these articles were written by looking at the fact that none of them published what else they did while on Facebook.

Just a few examples of what Facebook could (and probably does) track to adjust your news feed:

  • What articles you share
  • What articles you comment on
  • What status updates you make
  • What the textual content is of those comments, status updates, and shares and how that relates to the articles and other things you are (or may be) interested in
  • How long you linger on an update or parts of your timeline
  • What articles you click on from your news feed
  • What applications you use
  • What in-app purchases you’ve made
  • … and so many others

I will end it with my own piece of anecdotal evidence.  A few weeks ago, I had a really bad experience with Norwegian Airlines.  I posted a status update on Facebook about how bad it was, simply mentioning their name and where we were delayed.  Within a day or two, I started receiving advertisements on my news feed for Norwegian Airlines.  Was I their ideal candidate for advertising?  Probably not.  In fact, I’m probably happier with Wal-Mart, McDonalds, and Raytheon as companies than I am with Norwegian Airlines.  But it didn’t require me clicking “like” on anything from Norwegian to start getting them to show up in my feed… I just had to mention something they were targeting: flights to Scandinavia.

 

In short, you don’t need to go through and “like” anything to get the  “loathsome” or “banal” things in your news feed or affect your friends’ news feeds.  You just have to actively be on Facebook; it comes with the territory.

 

So go ahead and like, comment on, and share this post.  You’re not fooling anyone.

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *