Facebook penalising 'link posts', but by how much? About 240%

Potential Reach vs Reach on link post & non-link post | Create infographics

We've all heard that Facebook penalises some posts, right?

It's been clear for some time now that Facebook affords posts that encourage activity within Facebook more reach than those that drive users away from Facebook (ie - posts that link to your website), but what exactly is the difference in reach between a post that has a link that drives users off-site, and a call to action that keeps users chatting within Facebook?

According to a test I've conducted, it's about 240%

Now, this figure is the result of one, non-scientific test. I will be conducting more tests, but the stats are interesting and demand further exploration. The raw data from my test is available here.

Test set-up:

• Posted a post to Facebook twice, one including a link, the other a carbon copy of the first post minus the link.
• The linked post was targeted to Facebook Page fans aged 13 -34, a potential reach of 8,200. It was posted on March 14 at 11:58am and is available here
• The post without the link was targeted to Facebook Page fans aged 35 - 65+, a potential reach of 7,160. It was posted on March 14 at 12:00pm and is available here 

Test results:

• There was a massive difference in levels of reach and engagement between the posts.
• The linked post had a potential reach of 8,200, but only reached 2,128. The non-link post had a potential reach of 7,160 and reached 7,251 (The 100%+ reach is probably due to the level of likes, comments and shares resulting in the post reaching an even wider audience)
• The linked post generated one comment, 27 likes and three shares. The non-linked post generated 17 comments, 67 likes and 43 shares.

Test flaws:

As I said before, this test is not scientific, there are a number of reasons why my results/data might be skewed:

1. I've targeted two very distinct age groups. They may have very different typical levels of response to Facebook posts.
2. The post that I posted second came out on top - Does this have any impact on the stats (despite that fact that the posts were targeted to two different audiences)?
3. The post itself may have been a post that appeals more to an audience aged 35+.

Thoughts:

Despite all of the flaws with the test that I have pointed out above, the difference in reach, about 240%, is very, VERY significant. I need to conduct more tests before I come to any solid conclusions but the data above strongly suggests that Facebook heavily penalises posts that drive traffic away from Facebook.

You can help:

Do you have relevant data you can share? It would be great to get a broader overview of the impact of links on the reach of Facebook posts if you can share data from the Facebook Pages you manage.

Note: Obviously this requires further tests, and we should also test the impact of images and videos on reach and click-throughs. It isn't scientific, but it is very interesting.