Algorithm Experiment
Beginning on Saturday November 13 and concluding on Monday November 22, I conducted an Algorithm Experiment with members of my Facebook group Inspiration & Action for Young Leaders.
Background: I recently learned about the Facebook whistle blower scandal which revealed that Facebook incentivized angry content by prioritizing it in the algorithm over other types of engagement, starting in 2018. The outrage, reactions & comments that came as a result was called "meaningful social interaction".
Objective: I wanted to see if the "algorithm" in my facebook group produced a result where "anger" reactions were seen by more people and engaged with more. I was also examining who followed instructions, comments and the overall reach of the post.
Method - for 8 days I posted a version of the same request in the facebook group & asked for different emoji reactions each day. The post asked people to respond with one specific emotion only. I auto scheduled these posts to be posted at approximately the same time each morning at approximately 7:30am. The only outlier was the first post, the test post, where people could choose any reaction they wanted.
Take a look at the results poster!
Observations
On average, the daily participation was only 25% of group membership. That means 3/4 of the group is not engaging, at least not in the experiment.
I am surprised that "wow" was weakest reaction lol.
In classic facebook, style, "like" got the most amount of reactions. It is interesting that despite this,the amount of views was the same as the daily average.
Excluding the test post, the "love" post had the farthest emoji reach with "sad" coming in second place. Anger was last.
All the posts got 2 comments. Except for "love" & "care", those two were the only ones that generated no comments.
There were only 3 instances when people did not correctly react to the instruction. I don’t think this made a huge difference.
Conclusions
Daily Reach was 50 people on average seeing the experiment posts. Out of 200 group members and possible participants in this experiment, the posts reached 25% or one quarter of members.
Love Wins - the "love" post had the farthest emoji reach with "sad" coming in second place. Anger was last. These results were very surprising to me!
Mix It Up - since the test post did best, I think that when you have a variety of responses that those posts go farthest.
Time Limited - I noticed posts tend to stop receiving views after 24 hours. for that reason, I gave myself 48 hours to gather the data for our final emotion, anger.
A new question arises: has facebook now changed the algorithm to reduce anger and amplify love?
Was there anything you noticed that I didn’t mention? I am forever grateful to the folks who constantly support me and this time all my love and gratitude goes out to the folks who are members of Inspiration & Action for Young People for making this experiment so much fun. Until next time!
