You’ve likely seen many videos on Facebook the past year — both Live and uploaded — with massive amounts of views.
These views give people and page owners some level of credibility and “social proof” after all.
But I’m curious to find out if people are actually watching the videos, or just having them autoplay in the news feed.
The idea for this experiment came to me after doing a few Facebook Live videos on the Agorapulse page. I could see the views were pretty high, yet the interaction was just so-so.
Subscribe to the Social Media Lab Podcast via iTunes | Spotify | Google Play
As you can see Facebook says we had 1363 Video Views — awesome.
I’m just not sure how accurate this is, or how deceiving Facebook is being with this metric.
What are Facebook Video Views?
With four billion video views happening daily on Facebook what is considered a video view?
No one truly knows it seems. It’s similar on YouTube when it comes to view count.
Do you have to watch a certain percentage to count as a view? A certain amount of time?
We don’t know really.
Facebook has changed its reporting recently to show users a “10 Second Views” count. You’ll notice the video above had 557 10-second views, but 1363 Video views.
So which number is more important?
If your video is 45 minutes long who cares if someone only watched 10 seconds!
We know from a study by Wochit that video views were on the decline at the end of 2017 after Facebook announced news feed changes:
So with views on the decline, it’s important to understand more about the views Facebook is reporting to us, and then figuring out what to do with it.
Hypothesis
This study is a tad different than others as I’ll be pulling data from posts not necessarily intentionally published for this test.
This may be a good thing in the end as the videos were posted 100% naturally as any other page might.
What I want to figure out by the end of this experiment is are people actually watching our Facebook videos?
Hypothesis: The real number of viewers of Facebook videos is less than half of the views reported by Facebook.
How We’ll Test Facebook Video Viewership
For this test we’ll look at the video stats for the most recent 20 videos (a combination of Live and uploaded videos) for these 3 pages:
- Agorapulse – Our company Facebook page
- Geeks Life – This page is run by my old friend David Foster. He and his partners review tech products, talk about food, life, video games, movies, etc.
- Live Streaming Pros – Luria Petrucci (formerly Cali Lewis from GeekBeat.tv) runs this page along with David Foster. They pretty much only post Facebook Live videos, since they are the Live Streaming Pros after all! They have a few courses on doing live videos and install video studios for many of the social media gurus you follow.
Because Geeks Life and Live Streaming Pros focuses so much on videos, especially Facebook Live, I really wanted to dive into their Insights.
I’ll be compiling these 3 pieces of data for each of the 20 videos, and then averaging them together for each page and overall:
- Video Views
- Unique Views
- Percentage of users with Sound On
I wanted to compare the Unique Views to Video Views to eliminate anyone that might have watched the video more than once.
Then I wanted to look at that number and compare to the number of users that actually turned the sound on.
After all, if the sound is off does it matter?
Especially on a Facebook Live video.
Data on Facebook Video Viewership
Let’s first just look at each page’s average for the 3 data points we were looking at:
Agorapulse:
- Video Views: 324.95
- Unique Viewers: 307.70
- % with Sound On: 15.70%
Geeks Life:
- Video Views: 248.25
- Unique Viewers: 226.80
- % with Sound On: 27.75%
Live Streaming Pros:
- Video Views: 1131.05
- Unique Viewers: 1030.15
- % with Sound On: 40.60%
I also wanted to break down the numbers based on live and uploaded videos (all of Live Streaming Pros videos were live.):
Uploaded Videos to Facebook:
- Video Views: 203.58
- Unique Viewers: 190.77
- % with Sound On: 18.27% (Geeks Life- 23.0%, Agorapulse- 13.46%)
Facebook Live Videos:
- Video Views: 670.87
- Unique Viewers: 616.24
- % with Sound On: 32.30% (Geeks Life- 36.43%, Agorapulse- 19.86%)
To better get a handle on the numbers here are the averages of all 3 pages and both video types combined:
Average of all 3 Pages:
- Video Views: 568.08
- Unique Viewers: 521.55
- % with Sound On: 28.02%
Evaluating the Data on Facebook Video Viewership
All of those numbers are great, but what can we extrapolate from them?
Although it wasn’t a part of my original study I can easily see that nearly double the number of people watching Facebook Live videos turned the sound on 00 32.30% vs 18.27%. Which makes sense. (and also means only 32.30% of those watching the Live video are truly “watching”.)
I would assume people are listening to a Live video to interact with the person broadcasting, whereas they may not turn the sound on for an uploaded video that appears on the surface as a marketing video.
But the reason I did this study was to see if the video views Facebook gives page owners is accurate and reliable.
I’m gonna say no it’s not.
The number of video views seems inflated on every video and every page I looked at.
It seems Facebook is reporting video views regardless if a person watched it more than once, or if they even turn the sound on.
As the stats above stated for all 3 pages studied Facebook showed us 568.08 views, with 521.55 unique viewers. So, in reality, the average views are 521.55, who cares if someone watches a video more than once.
Even that number is false IMHO though.
On average only 28.02% of users that “viewed” the videos turned the sound on. Which equates to just 146.14 people.
That’s the more accurate video view count to me. Unless your video was just a short video with text flashing on the screen having the sound on is crucial.
It could be that Facebook just assumes the audio isn’t always important, but imagine watching the news with the sound off.. It would have no value and wouldn’t make sense. And those with the sound off likely aren’t going to buy from you.
Conclusion about Facebook Video Views
My original hypothesis was the real amount of users viewing a video on Facebook was less than 50% of what Facebook reports as video views.
According to the data I pulled from these 3 pages that’s 100% true.
I am concluding that with only 28.02% of users turning the sound on that’s the true number of viewers.
This especially true when it comes to Live videos that require listening to know what is going on, which in our study was 32.30% of users turning the sound on for Live videos.
So don’t get crazily excited if you get 1000 views on a video, if only 10% turned on the sound that means 900 of those that “viewed” your video weren’t really viewing it and actively a part of your video.
They just likely saw it autoplay in the newsfeed.
That’s the other stat that Facebook doesn’t share with us — how many users simply had the video autoplay.
I personally have autoplay turned off for my desktop and mobile experience on Facebook, it’s annoying and I’m on slow internet.
To get a more accurate idea of the number of viewers I would advise looking at how many turned the sound on and looking at the graph of how many completed the video:
I got to this screen by clicking on the video in the page’s Insights.
Click on “Video Average Watch Time” to get to the “Audience Retention” graph:
Facebook lets you hover over the blue line seen there to see the percentage of viewers that watched at different points within the video.
You can also see on this view the number of those that auto-played or clicked to play.
But this is also another stat that is hard to get a grasp on. This video shows 92% auto-played yet 21% had the sound on.
So either those that auto-played also had sound on or some made the choice to turn the sound on. Again it’s kinda confusing.
For me knowing how many had the sound on is the most important stat Facebook gives us on our videos (clicked to play is nice to know as well).
Without the sound on I won’t count them as viewers, and you probably shouldn’t either.
Thoughts?
