Last month, EdgeRank Checker posted the results of a study they did that showed an apparent discrepancy on Facebook when scheduling posts via third party applications like HootSuite, TweetDeck, Seesmic and others. While I don’t doubt the metrics that EdgeRank Checker used, nor did I doubt their findings, we decided it would behoove us to see if the same results were true here at Paramore. Turns out they weren’t quite as one-sided as EdgeRank Checker had found when it came to our clients. We ran an A/B Test between two clients to see if posting from HootSuite really affected our engagement rates and these are the results.
Summary: We were looking to see if the study about third party apps on Facebook not getting good results in engagement really held any weight with our clients. The short answer is “No.” There was really nothing statistically significant about the differences to imply that posting from HootSuite or Facebook is better for Comments, Likes and Feedback Percentages. There’s an entire spreadsheet of data with all the numbers and everything I looked at, but I’ll run down some of the main points here for you.
Methodology: I documented posts made on the Facebook Pages for Client A and Client B for two weeks and noted whether I made the posts from HootSuite or Facebook. I tried to get as even a distribution as possible for both posting sources as well as the time of day of the posting. HootSuite ended up with a few more posts because of late-night scheduled posts and weekends, but the differences aren’t significant and since I calculated averages, it didn’t make a difference. Once all the posts were made and the analytics were released from Facebook, I graphed the averages for posts made from HootSuite for Feedback Percentage, Post Comments and Post Likes and then did the same for posts made from Facebook for both Pages. I also looked at the data from posts made in the morning as compared to posts made in the afternoon/evening and then did another comparison between posts made with links attached and those without links (those updates made without links all had pictures attached instead of links).
Note: For whatever reason, on posts made Wednesday the 21st for both clients Impressions were low all through the day giving high Feedback Percentages. Those posts I have highlighted in yellow on the results charts. When those numbers were taken out, the averages were even more similar (as reflected in the numbers presented in the results summaries).
Results:
For Client A, the average Feedback Percentage on posts made from HootSuite was 0.21 and the average from posts made from Facebook was 0.19. The average for HootSuite Post Likes was 14.8 and for Facebook it was 13.5 and the average for HootSuite Post Comments was 3.26 and for Facebook it was 2.45.
For Client B, the average Feedback Percentage on posts made from HootSuite was 0.18 and the average from posts made from Facebook was 0.16. The average for Post Likes was 7 for post HootSuite and Facebook and the average for HootSuite Post Comments was 4.92 and for Facebook it was 6.44.
For Client A, the average Feedback Percentage for posts made in the morning was 0.22 and the average for posts made in the PM was 0.18. The average Post Likes in the mornings was 12.2 and for posts made in the afternoons/evenings was 15.3 and the average for Post Comments in the mornings was 3.93 as compared to 2.36 in the afternoons/evenings.
For Client B, the average Feedback Percentage for posts made in the morning was 0.14 and the average for posts made after noon was 0.2. The average Post Likes in the mornings was 6.1 and for posts made in the afternoons/evenings was 7.9 and the average for Post Comments in the mornings was 4.72 and was 6.5 in the afternoons/evenings.
For Client A, the average Feedback Percentage on posts made with a link attached was 0.22 and the average for those posts without a link and only a picture was 0.15. The average Post Likes for “linked” posts was 14 and the average for those without a link was 15.6 and the Post Comments average was 2.76 for “linked” posts and was 3.6 for posts without links.
For Client B, the average Feedback Percentage on posts made with a link attached was 0.18 and the average for those posts without a link and only a picture was 0.14. The average Post Likes for “linked” posts was 7.29 and the average for those without a link was 6.29 and the Post Comments average was 4.14 for “linked” posts and was 8.43 for posts without links.
Thoughts/Learnings:
There’s no significant data that says one way or the other if posting from HootSuite or Facebook is better for engagement rates for our particular clients under our circumstances. There are some numbers that are larger than others, but not to a degree that is statistically significant either way.
There’s a large discrepancy in Client B’s Comments on posts without links compared to posts with links, but that can attributed to a promotion running during the testing period that rewarded comments made on posts with pictures.
It appeared that we got more comments and Likes on Client B’s Page in the afternoons but when it came to Client A, the numbers were split (more Likes in the afternoon but more comments in the morning)
When it comes to our specific clients, there isn’t any reason why posting from HootSuite over Facebook would be any better or worse than the other. While our A/B Test doesn’t by any means disprove the one done by EdgeRank Checker, it does show the importance of understanding what works for your unique situation.
Have you crunched the numbers? What do you find works best: posting from third party apps or directly from Facebook?
Share Article
0 comments