Facebook Engagement: We’re Still Doing It Wrong

We’ve been having this debate internally for some time, but Toby Margetts' recent post on “Why You Are Measuring Facebook Engagement Inaccurately” has prompted us to get the pen out to reveal our latest thoughts on Facebook engagement.

We’ve been having this debate internally for some time, but Toby Margetts’ recent post on “Why You Are Measuring Facebook Engagement Inaccurately” has prompted us to get the pen out to reveal our latest thoughts on Facebook engagement.

There are a number of schools of thought emerging on this debate, and in essence, the split is largely coming down to those who believe in reach versus those who believe in measuring against total fan count (page audience size).  Personally, I sit on the total fan count size, as measuring based on reach is defying the purpose of Facebook’s EdgeRank. If your reach is low (because your overall engagement is poor), using this to show your engagement percentages as a higher figure is incorrect. You should recognize the low engagement and focus on increasing this so that your reach increases.

The problem, put simply though, is that there is currently no industry standard way to measure engagement. Why does this matter? Facebook’s EdgeRank algorithm determines the reach of your posts, so efficiently measuring and optimizing for engagement is key to extending your pages’ reach. 

An Engagement Metric Comparison

The table below is based on the same data on the same Facebook page. This page is performing particularly well for its sector and against local competitors. We ran the last week’s worth of data through the different engagement calculations we were aware of to show the vast differences shown:

Method

Engagement Rate

Talking About This (Audience)

31.4%

Talking About This (Minus New Fans, Audience)

14.6%

Daily Page Engagement (Audience)

6.1%

Daily Page Engagement (Reach)

4.7%

Average Post Engagement (Reach)

4.3%

Average Post Engagement (Audience Size)

2.7%

Let’s briefly go through the pros and cons of each metric:

  • People talking about this: Using Facebook’s open graph to find the PTAT number of a page and dividing this by total fans of the page (PTAT/total audience) is inaccurate as it includes all types of engagements, including new fans, so it is therefore influenced by any fan-acquisition campaigns that are occurring, clouding the true engagement rates of posts.
  • PTAT (minus new fans): Taking the new fans out of the talking about this number  — (PTAT – new fans)/total audience at time makes this a fairer metric.  The other advantage of PTAT is that it is publicly available data, meaning you can measure and compare against any page based on this, allowing for competitor analysis. However, optimizing for PTAT can be difficult, as the figure is always presented as a moving weekly figure, making post-level insights difficult.
  • Daily page engagement (audience): Daily page engagement (likes + comments + shares on a given day/fans on a given day) seems to be the middle ground between PTAT and average post engagement, and it certainly takes the middle ground in terms of actual engagement numbers in our example. From my perspective, daily page engagement is fundamentally flawed – to go to the effort of calculating the engagement daily is just as much effort as calculating per post, but without the level of insight that gives.
  • Daily page engagement (reach): As described above, daily page engagement is not a worthwhile metric in my opinion. On this version, the difference is between reach and audience, with the formula for calculation being likes + comments+ shares on a given day/reach on a given day. The issue with reach-based engagement calculations is that you can only see this for pages you own, meaning accurate benchmarking is impossible, and they don’t truly show engagement in my opinion.
  • Average post engagement (reach): Average post engagement gives you more insight, as you can break this down to show engagement levels by individual post to allow you to see what is and isn’t working. The average engagement of all posts: (comments + likes + shares for each post)/post reach. However, the major flaws are that this is a time-consuming process and that you can only get post reach for your own page.
  • Average post engagement (audience size): Average post engagement based on audience size: comments + likes + shares for each post/page fans. This suffers the same issues as the above metric but to a lesser extent.  You can’t get an accurate number on the audience size of competitors easily, especially if it’s more than four weeks ago. Tools such as Socialbakers can help with this problem. Alternatively, you can calculate this for a four-week period (and track it over time) by using the new likes per week metric on the PTAT graph and taking the new likes away from the total audience shown to give a fairly accurate estimate.

Weighted Engagement

In principle, weighting the different types of Facebook engagements (comments, likes, and shares) based on their importance is valid. However, using this as an engagement rate percentage is statistically incorrect.