Facebook “like” count 39% accurate

The number of “likes” usually displayed alongside the Facebook like button is really an aggregate of shares, likes and comments. This morning, I took an arbitrary mix of Facebook related stories and found that the actual number of likes only accounted for a 39% of the number displayed. This is by no means scientific, but I think it’s noteworthy.

Why does this matter? Because news sites are publishing factual inaccuracies in articles that say things like “100 people recommend this” when in fact only 39 people did.

Here are the stories that I looked at this morning:

All told, the like buttons claimed that those pages were liked or recommended 4,622 times. In fact, they were liked or recommended only 1,790 times.

In case you didn’t click on the first link in this post, I got the real “like” numbers by using my RealShare tool.

Feel free to take a look at the data in my malformed Google Doc.

Page views don’t measure audience

Recently, a colleague mentioned that he thinks increasing page views is different from growing an audience. He’s right. Here are a few definitions of the phrase “page view” and the word audience.

Audience: One Merriam-Webster definition of audience is “a reading, viewing, or listening public.” Dictionary.com defines audience as “the persons reached by a book, radio or television broadcast, etc.”

Page view: Google defines a page view as “an instance of a page being loaded by a browser.” The Wikipedia definition is “a request to load a single HTML file (‘page’) of an Internet site.”

So page views measure how many times a page has been requested or loaded, regardless of how many many people (an audience) made those requests.

Imagine each reader as a diner in a restaurant, and each page as a menu item. One person could visit three times a day and consume three hearty investigative reports. She is still one diner. Maybe another patron really likes tasty pictures, and orders 20 of them in one sitting. He is still one person.

But that restaurant analogy only represents one part of a reading audience. What if one person reads three of your articles on his tablet, skims two of your blog posts on his mobile phone, glances over one of your tweets on his work computer and enjoys three of your pictures on his personal laptop. Although he has looked at nine pieces of content – three of which aren’t traditional “pages” – he is still one person.

So if not page views, what do you use to measure your audience? Leave your comments below.

Oh, and make sure to read these important posts about page views as a metric.

Image courtesy of Stuart Mudie via Flickr.

Facebook button count is wrong, use RealShare

That number next to the Facebook share, like, or recommend button on most blog posts and articles is just plain wrong, so I’ve created the RealShare bookmarklet to show the real Facebook statistics for any web page.


Drag the RealShare button below to your browser’s favorites/bookmarks bar (scroll down for a video showing how to do this), then visit any web page and click on the RealShare button to see the actual number of people who have shared, recommended, or commented on that page.


What’s wrong?

Facebook’s Like Button plugin instructions provides an option to display the “total number of likes to the right of the button” but the documentation for their Share button explains that they really aggregate three numbers: “the total number of times the page was shared on Facebook, how many comments were added to the story shared on Facebook, and how many times friends Liked the shared story.”

In the screenshot below, it appears as if this story about WoW addiction was shared 313 times. The RealShare button reveals that it was actually shared only 138 times. The inflated number reflects those shares along with the 35 times the post was liked/recommended and the 140 comments left about that page.


What to do

Does the misleading number in that Facebook button put you at risk of losing credibility with your readers? Every reputable news provider (including independent bloggers) should answer that question. I’ve just disabled the Facebook button on this blog, and I hope you consider doing the same until Facebook fixes this accuracy and credibility problem.

Installation video

Thank you to Ian Hamilton (@hmltn) for asking the question, Clint Watson (@clintavo) for reporting the news back in July, and for All Facebook for its Facebook Like Count Calculator that inspired me to create this little tool.

Measure engagement with one click

I created this BackType bookmarklet to quickly measure the reach of articles and blog posts in social spaces.

Drag this button BackType Stats to your browser’s bookmarks tool bar. If you are using Explorer, make sure your “Links Toolbar” is showing, then right click on BackType Stats, click “Add to Favorites” and save the link in your “Links” folder.

Now, when you’re viewing any web page, just click on the “BackType Stats” button to see how that page is being shared and discussed on Twitter, Facebook and reddit.

I won’t get into the details of BackType, but it’s a service that shows at least these details about any web page:

  • How many times a link to the page has been tweeted
  • How many times those links have been clicked
  • How many times the link has been shared on Facebook
  • How many times the link has been liked on Facebook
  • How comments have been posted on the page
  • How many times the link has been commented on in Facebook
  • How many times the link has been commented on in reddit

Take a look at BackType’s analysis of a post about the Knight News Challenge by The Nieman Journalism Lab.

BackType example: Knight Challenge