Engagement Rate - out of Fanbase or out of Reach?
Facebook have started rolling out the new Facebook Insight dashboard to Page owners and I have to say - it's pretty awesome! While taking notes to review it, I noticed that Facebook have introduced a new metric - their own version of Engagement Rate, to replace Virality for Page posts.
This is Facebook's definition of its Engagement Rate:
Engagement rate is the percentage of people who liked, shared, clicked or commented on a post after having seen it.
This is an important move for Facebook as it distances itself from Virality (which was briefly available along with the Engagement Rate, until it got pulled on Friday).
On the subject of engagement, there are two ways that marketers get their engagement rate: out of fanbase and out of reach. The difference is not specified in Social Monitoring tools, while only a handful are transparent enough to show you how this metric is being calculated.
But what if I told you that the Engagement Rate you've been reporting on is actually not what you're looking for?
Engagement out of Fanbase
Engagement out of fanbase is a metric that calculates the ratio of engagement out of the number of fans. It's the most known approach, the default formula for quite a few analytics tools out there, and the method that most marketers swear by.
It follows the following formula:
[sum of engagement] / [total number of fans]
Some tools and agencies define the sum of engagement as the total number of main engaging actions that users can perform (i.e. likes, comments, and shares). However, the most common approach is to take the PTA(T) [People Talking About (This)] as the 'sum of engagement'. By using this approach, you're counting in the following actions:
- liking a post
- liking a comment
- commenting on a post
- commenting on a comment
- sharing a post
- answering a question/"poll"
- responding to an event
- claiming an offer
- posting on the page wall
- mentioning the page in a post
- tagging the page in a photo
- checking in at your place/page
- sharing a check-in deal
- liking a check-in deal
- writing a recommendation
- liking a page
Engagement out of fanbase has been used as an indication of how much of your fanbase you're able to engage with your content. The most notable advocate for this is SocialBakers, who came out in March 2012 with a viral blog post on how to calculate engagement rate on Facebook and Twitter.
Engagement out of Reach
Advocates: Facebook, SimplyMeasured
Engagement out of reach is a metric that calculates the ratio of engagement out of the number of people reached. It's a less known approach as most tools don't actually offer this, and it's not even an approach that is being talked about most bloggers and marketers either. However, it's interesting to see that Facebook favours this approach in its own Insight dashboard.
The way this metric is calculated is by using this formula:
[sum of engaged users] / [people reached]
So, which one's better?
First of all, the real question is: what do you think the Engagement Rate is? Let's settle on a definition of Engagement Rate. I propose the following:
Engagement Rate is the measure of how well your audience responds to your Page content.
Agree? Ok. Now, if that is what you're looking for in your Engagement Rate metric, calculating the Engagement Rate out of Fanbase is not what you're looking for, and it's a mathematically incorrect metric. Here are the 5 main reasons why:
1. Using the PTA(T) as your numerator is an unfair representation of 'engagement' Let's look back at what PTA(T) is: it's the sum of all possible interactions between a Facebook user and your Page (along with its content) as long as those interactions create a 'story' (more on this later). Hence, if 100 people click 'like' on your Facebook Page, 100 will be added to your PTAT. However, what if only 5 of the 100 people who liked your Facebook Page today actually interacted with your content? Then PTAT will add those 5 interactions to the previous 100 likes, leaving you with an inflated number that does not correctly represent interactions with your content.
2. Not all interactions are included in PTA(T) Clicking on your post, opening a picture on your page, pressing play on a video, and clicking on a link are just some of the ways Facebook users can interact with your Page content. However, those won't be counted towards PTA(T). Why? Because they don't create what Facebook calls "stories". A story is an item that is displayed in your News Feed or News Ticker (until the old layout completely fades out). So, an example of a story is "Ben liked a picture", "Ben commented on a post", "Ben was tagged in a picture". However, clicking on a page post, opening a picture, pressing play on a video or clicking on a link does not create a story. Because of that, those interactions aren't counted in PTA(T). However, those actions are still a part of engagement - a Facebook user might not click 'like' on a video you've posted on your Page, yet if he clicks on play and watches it, he still engaged with your content. Using PTA(T) as your numerator means that all of these interactions that do not explicitly create a story will get lost - even though they're an equally important chunk of possible interactions and engagement between your fans (and non-fans) and your Page.
3. You're assuming that only your fans are engaging with your content Using the number of fans as your denominator is a wrong assumption for one big reason: your audience doesn't consist of just your fans. In fact, you might even see that a big chunk of people who engage with your content are not even your fans (to which you can then ask - why is my content encouraging people to engage with my Page and not be a fan of it?). This wrong assumption will also result in a wrong calculation. By doing so, the interactions by users who aren't your fans will be wrongly attributed to those who are your fans. Simply put, you're using a ratio of interactions made by fans AND non-fans out of the number of fans. The only way this formula could be accurate is if you only counted the number of interactions made by your fans.
4. Your audience doesn't only consist of your fans There seems to be an ever-increasing obsession on the number of fans - getting more fans, increasing the fanbase... This is also reflected in the fact that the denominator for the Engagement Rate out of Fanbase is the number of fans on a given day. However, the reality is - when it comes to the engagement on your Page, what really matters is the audience you had at the time of posting and how many out of that interacted with such content - regardless of whether they're your fans or not. The metric that can give you the size of the audience isn't the total number of fans - it's the Reach.
5. If you use PTA(T) as part of the numerator, you're not counting people but interactions Since PTA(T) stands for 'people talking about this', it's easy to think that if you divide it by the number of fans you get the rate of engaged users for that content. However, despite the name, PTA(T) doesn't count people who engage with your content, but their interactions. One fan can be responsible for 20 likes, 5 comments and 3 shares, and it would be counted as "28 people talking about this".
Still not convinced?
Here's a very practical and straightforward example to describe this:
160 people RSVP 'yes' to an event. However, on the day of the event, only 80 people attend, while the other 80 are still marked as RVSP'd 'yes'. Throughout the event, 20 people engage with the speaker, be it by asking questions or joining in the conversation, while the other 60 just listen intently.
Here's the question: if you were the speaker, how would you calculate the ratio of engaged people, i.e. the engagement rate of the event?
Now let's just say, for argument sake, that there are 10 +1s in the audience. These are friends of some of the people in the audience who had RSVP'd 'yes' to the invitation, even though these +1s aren't actually registered. Out of the 10 +1s, only 4 engaged with the speaker.
Now, how would you calculate the engagement ratio? Would it look something like the following?
Now, ask yourself - why didn't you calculate the rates like this...
Perhaps you chose the first method because, after all, if you want to calculate the rate of engagement of the registered attendees you need to count in those who attended, not everyone who RSVP'd 'yes' - it wouldn't be fair to count in the people who were not part of the audience: the very fact that they were not part of the audience means that they wouldn't have been able to engage with the event in the first place.
Perhaps you chose the first method because, after all, if you want to calculate the rate of engagement of the +1s, you should only factor in the number of engaged +1s and their total number. After all, why would you divide the number of engaged +1s by the number of registered users, when the +1s aren't even registered?
Perhaps you chose the first method because, after all, if you want to calculate the overall engagement rate, you need to count in the engaged registered users as well as the engaged +1s (regardless of the fact that they weren't registered), and divide that by the total number of people in the audience - the registered users and +1s.
If you agree with the three statements above, then you understand the principles behind calculating the Engagement Rate out of Reach rather than out of Fanbase.
So what's the 'magic formula'?
First of all, let's recap what we need from an engagement formula, before we come up with one.
- The Engagement Rate should include all interactions, whether they create a Facebook story or not. This includes viewing photos, clicking on links, playing a video etc.
- The Engagement Rate should not exclude users who aren't fans of your page, especially since they too can interact with your content without being your fan.
- The Engagement Rate should be taken out of the audience of your content, hence your Reach. Only this number can give you a realistic indication of who saw your content, thus having the chance to engage with it.
Based on those 3 principles, here's how you would calculate the Engagement Rate out of Reach:
[Engaged Users] / [Reach] %
Even though we've been discussing engagement rate for your content, you can still use the same principle to derive a formula for the engagement rate for your pages. This is especially useful if you need to benchmark your Page against other pages you own. Just use this formula:
([Engaged Users] / [# of posts]) / [Reach]
If you do have a tool that calculates engagement out of fanbase, or you're not sure how the engagement is being calculated in your tool, I recommend you start a discussion with the company that provided you the tool. You could even send them this article to initiate a discussion on how engagement is calculated and changes that should happen in the tool thereafter.
The good news is, whether your tool calculates engagement out of reach or fanbase, or even if you currently do not have a tool, you can STILL calculate this on your own by using the Engagement Rate formulas.
Then, if you want, you can even break that engagement rate down in 3:
- Engagement Rate out of Organic Reach: [sum of engaged users reached organically] / [organic reach] - Engagement Rate out of Viral Reach: [sum of engaged users reached virally] / [viral reach] (to see the propensity of your fans' friends to engage with your content) - Engagement Rate out of Paid Reach: [sum of engaged users reached through paid content] / [paid reach] (to justify paid campaigns on Facebook - if you've ever needed a metric to justify getting more budget in your social campaigns, this is the metric you'll need)
One last note...
No two tools are the same. In fact, you can have two tools measuring a metric that is called the same across the two, but is calculated differently.
For example: according to SimplyMeasured, engagement is likes, comments and shares; however, according to Facebook, it's likes, comments and shares, as well as clicks on the post (e.g. opening a picture, clicking on 'play' on a video, clicking through a link).
Here's the lesson: question everything about your analytics tool - whether it's a free tool or a premium one, whether you've had it for a long time, just acquired, or you're thinking of adopting one. Make sure you understand exactly how each metric that you're going to report on is calculated, and try and match it with your own calculations to see how accurate your reports are going to be if you decide to rely on whatever tool you have.
Let me know in the comments section what your thoughts and don't forget to share this article!