EdgeRank Algorithm DOES NOT Penalize Publishing Tools (we have the proof)

Let’s start from here:

Barbara Gozzi
@postpickr did you spot any problems about the visibility of the posts scheduled by PostPickr on #facebook? It’s probably too soon #testing mode on

This tweet sent by a beta tester of ours was chosen as the proof that a certain grade of skepticism on the external publishing tools (like HootSuite, Buffer and also PostPickr) is still current among the experts. We have already talked about in this post: although its heading ends with a question mark, the whole following consideration does clear all the doubts. Some days ago we hit on the umpteenth article in which the skepticism is restated:

[…] Not to mention other types of post, i.e. the ones posted automatically through some tools like IFTTT or other ones. With the recent roll-out of EdgeRank, Facebook has reduced their visibility much more.

In the following discussion we raised objections and reported the consideration included in our previous post, but the author answered us in this way:

I regrettably don’t agree with you anyway. My opinion is based on my personal experience. I’ve been managing some facebook pages for a long time and I could certify that the posts published automatically through IFTTT are actually penalized because of their reduced visibility. My statement is based on a concrete comparison between the posts I published myself and the automatic posts. The result highlights a remarkable difference between the two of them.

Although IFTTT is not a proper publishing tool like PostPickr, his answer persuaded us to conduct an internal test with pure scientific approach. The variables that potentially could falsify this kind of test are actually many and according to us, apart from the kind of content and the modality of post publication, it is not enough to compare a “native” post with a post published through an external tool.

The Sample

For this test we chose a Facebook Fan Page that we have been managing for some years (since its birth in 2009) and we perfectly know the trends of. It is the rock band Le Rivoltelle fan page. Its subscribed fans were over 13k at the moment of the test and each of them signed up for the page spontaneously (we the managers of the page have never had recourse to sponsored ads). We would like to underline this last feature because it seems to us an important quality index of the sample analyzed here.

The Contents

We chose a kind of content which could generate the same level of interest in relation to the audience of the page. This was a crucial step for a correct evaluation of the results: we rejected the idea of using an image post type or a post with shared links, because their peculiarity (quality of the image, link formatting, reliability of sources, narratives, etc.) also may influence significantly on the user engagement (who in turn influences the Organic Reach). We needed the most neutral and repeatable type of content (and form as well).

We were lucky: in this Facebook Page we always say hello to everybody in the morning through messages wishing a good day to all fans in different combinations. We decided to use always the same publication format for the test:

 

Good morning! :)

Frequency, Modality and Date Range of Publications

The kind of content chosen for the test let us opt for a daily frequency of post publication, so that we could remove the variables depending on the single days of the week. The same thing was done with publication time (9.30 a.m.) that we consciously did not respect each time for just some minutes in order not to make it look like there was some automatism. The posts were published continuously in the period between 18th September and 3rd October 2014, alternating Facebook and Postpickr, except for two sessions in which we published posts for two days through the first or the second method. In the two Sundays included in the test period of time we used this format:

 

Have a good Sunday everybody!

Goals, Measurement and Results

The main goal of the test is to establish whether and how much the source of publication influences the Organic Reach, that is the number of people to whom a particular kind of content has been showed.

The measurement was conducted 4 days after the publication, a period of time that we found suitable to get a stable enough result.

So here is the results:

Analisi comparativa della reach tra i post pubblicati con Facebook e quelli pubblicati con Postpickr

Among 16 posts in all, 8 of them published through Facebook have achieved 14.445 in terms of Reach, with an average of 1.805 visualizations per post. The 8 posts published through PostPickr have reached 15.300 visualizations, with an average of 1.916 visualizations. Also we have measured the quantity of received likes: Facebook posts have reached 82 likes with an average of 10 likes per post, whereas Postpickr posts have reached 104 likes with an average of 13 likes per post. By the way, it must be said that the number of the received comments reflects this trend as well.

Conclusions

Although Postpickr took a little advantage on Facebook (we cannot be but satisfied for this! :D), we do say that there is no significant variation in the results. The trends are absolutely comparable and, in some cases, unnoticeable. You can verify everything by looking at this analytical table in which this study is provided with documentary evidence: you will find datas, measurements, links of sources and screenshots.

We are waiting for your opinions, especially about the methodology we have chosen to use: did it look convincing to you? Let us know your opinion in the comments!

(Many thanks to Vincenzo Lotito for the English translation of this post!)