After numerous disputes on the Internet and seeing no single point of view, we decided to thoroughly investigate the issue of how audience engagement rate (degree or level) should be calculated, which formula is the most correct and why.
To start with, let’s discuss the basic definitions, which may be useful and are often confused with Engagement Rate:
Engagement is a unit of audience response: like [favorite], comment, share, sometimes clicks (following a link), reactions for Facebook and dislikes for YouTube, or retweets
Engagement Volume is the total (sum) of engagements
Love Rate is the ratio of likes count to audience size (total of all likes/count of followers *100 %)
Talk Rate is the ratio of comments count to audience size (total of all comments/count followers * 100 %)
When calculating for an average post, the count of posts is added to the denominator
Reach is the count of people who contacted the post (or ad) at least once
Views is the post’s count of views
Engagement Rate Calculation Methods in
There is no single opinion as well. However, some basic formulas and definitions tend to prevail:
Engagement rate is a metric that measures the level (degree) of engagement received from the audience by a specific part of the published content. *- https://trackmaven.com/marketing- dictionary/engagement-rate/
Social media engagement rate can be of three main types: daily engagement rate (ER day), engagement rate by posts (ER post) and engagement rate by reach (ERR). It’s worth noting that international services and companies distinguish the indicators by type very seldom: ER post may be called the classical indicator in one context, and in another context it might be ERR. Therefore, when working on the Popsters content analytics tool, we initially distinguished those metrics to avoid confusion.
** - http://simplymeasured.com/blog/facebook-metrics-defined-engagement-rate/
Types and Descriptions of Various Engagement
Engagement Rate by Reach
The basic engagement rate formula for a single post is:
The resulting value will show the % of people who were active with a specific post.
Average ERR will show the % of people who showed interest to page’s posts among those who saw them.
In most cases, social media do not distinguish between reach to followers and total reach. In such case, a post with enough viral potential (viral post), having reached outside the community due to shares and having appeared in the feed of their friends and followers, would receive far reach and relatively smaller engagement (with respect to the total reach value). This results in decreased ERR percentage of the post.
In another situation, which is particularly true for social media with algorithmic feed, the post may appear in the feed of a very narrow fraction of the audience (which is especially noticeable for small pages) – just a couple of dozens of persons who are likely to show activity since they are interested in that page’s content in the first place. The very low reach will then provide a relatively high engagement level.
As a result, the ERR comparison between a viral post and a non-viral post shows substantially higher engagement indicator for the latter.
It is quite simple to verify those conclusions – you can download post statistics with reach, calculate the ERR and sort by it. This will result in objectively popular posts ending up far from the top.
Conclusion: Engagement rate by reach will suit a situation where you need to evaluate the approximate % of people showing activity among those who viewed the post, but it is totally unsuitable to comparing quality or popularity of those posts and it will be biased when different pages or their posts are compared.
Daily Engagement Rate (ER day)
“Supposedly” is because this metric’s calculation conditions do not account for quite a big error: as for other ER indicators, every active follower could show activity several times, but for ERday this error becomes more noticeable as activity by one person on different posts published in the same day will affect the final value more.
Another drawback of this metric is that posting frequency is not accounted for – if you want to improve ERday, just post more (while the condition is filled that the more there are posts per day, the higher is the total activity for them).
Conclusion: Daily engagement rate will help to understand the approximate activity % in relation to the number of followers, in other words, the effective reach (engagement volume) of a page, but it can be hardly considered more useful than the net engagement volume when expressed as %.
Engagement Rate by Posts (ER post)
The basic formula is shown below:
The resulting value shows the [average] % of people engaged by a specific [or average] post.
Unlike ERR, this indicator is based on an absolute value in the denominator – the count of followers, and it will not vary dramatically from post to post. This will bring quality and interesting posts to the top when sorted.
The drawback of this indicator is that when the count of followers increases, the total engagement percentage decreases [in relation to the total number of followers]. There are two reasons for that: the old followers’ interest to the page may decease with time as they add more communities to their feed, therefore posts of the page catch their eyes less frequently, etc. This is why pages with drastically different audience sizes cannot be compared by ER without correction – average indicators for the pages should be taken into account before comparing.
Conclusion: ER post will help to assess how interesting are posts to the audience, compare posts and activities in different communities (with correction if the audience sizes differ significantly).
Other Approaches to Engagement Rate Calculation
Some companies and services use other approaches to ER calculation:
Engagement Rate by Views
The calculation is similar to that for ERR but, instead of reach, count of post’s views is used:
ER view = (sum of post’s engagements)/(count of views) * 100 %
Average ER view = (sum of ER view of all posts for the analyzed period) / (count of posts for the analyzed period)
This approach to the indicator has the same disadvantages as ERR but in addition there is another one: the views indicator accounts for non-unique views by the audience, i.e. multiple views by each viewer are added up when the post is opened repeatedly, which can cause even larger difference between indicators for different posts than with ERR. These are the reasons we wouldn’t recommend to use ERview for any qualification of community contents.
As one can note, in all the previous approaches ER is a percentage indicator based on audience size (or reach). Thus, one may say that:
Engagement rate is a metric reflecting the % of audience engaging in community’s event.
As a result, factored engagement cannot be called ER, since for factors other than 1 the resulting figure will not show the % of the engaged audience. Such metrics might be called like “activity index” or “virality index”, depending on the factors used, but certainly not engagement rate.
This metric is almost identical to ER post calculation but the multiplier, in the end, is not 100 % but just 100. As a result, the metric shows not “% of the audience engaged by post” but “average number of reactions per 100 followers”. This definition negates the drawback common for percentage ER indicators – the possibility of multiple reactions from the same followers, as there is no reference to the audience size.
This new definition is the problem of such approach: the classical interpretation refers to the audience, the size of which must be also taken into account, and the new format does not have it, which creates an impression that this indicator is an absolute value and can be compared between various pages without taking the audience size into account.
Engagement Rate by reach cannot be deemed objective for assessment of post’s quality or comparison between different pages, as it is a relative indicator. In our view, it is more correct to compare the reaches themselves or the reach to followers count ratios, and then you will see the approximate % of the audience reached by the page’s posts.
ER day will be useful to assess the engaged % of the audience on the whole, but remember this indicator’s drawbacks and that it is certainly not suitable for content qualification. The most practically useful metric is ER post – it will help to assess and compare popularity for different posts, compare (with corrections for various audience properties) different pages.
Accordingly, in case the objective of community management is the development of audience engagement, it is reasonable to check this indicator's dynamics for the page for different periods. Engagement rate formulas in this article are applicable for Instagram, Facebook, Twitter, as well as for other social networks.
It's also worth noting two more indicators, LR (Love Rate) and TR (Talk Rate). Their calculation formulas are almost identical to that of ER post, but the first metric focuses on “love” shown to the page’s posts and the second one – on followers’ talkativeness. These indicators allow comparing pages by their posts’ degree of attractiveness or the audience’s level of communicativeness.
Useful article on a similar topic: