Photos draw greater rates of engagement than any other type content posted in Facebook, according to a recent study conducted by Web Liquid.
The statistics are interesting in that they confirm what most marketers have known for a while — more dynamic content leads to greater levels of engagement. However, it’s not safe to assume these statistics can be applied to all pages.
The digital marketing agency reported the following engagement rates for the four types of content most commonly posted in Facebook pages:
- Photos: 0.37 percent
- Videos: 0.31 percent
- Text: 0.27 percent
- Links: 0.15 percent
These numbers represent a general reaction to content, but they’re sure to differ from page to page. The only way for an organization to measure the engagement levels of content by type is to actually go out and measure it.
Here are five simples steps to replicating the Web Liquid study with data of your own:
1. Export ‘Post Level’ Data.
The online data Facebook displays is great, but the good stuff is in the download. Export three months worth of data from the Facebook Insights tab. Be sure to specify “Post-Level Data.”
Once the Excel file is downloaded, work within the second tab across the bottom, “Lifetime Talking About This.” This is the tab that lists each post, the date it was published and the number of times it was liked, commented on, or shared.
2. Create a Column for Content Types.
The online Insights interface is great because it places an icon next to line of post data to define the content type — status update, photo, video and link. Unfortunately, the downloaded data doesn’t differentiate so an additional column needs to be added.
Simply insert a column after the Message and title it ‘Content Type.’ Code each line accordingly as one of Facebook’s four types of content. This is the most labor-intensive part of the process but even it doesn’t take much time.
3. Add Impressions.
The “Lifetime Talking About This” tab doesn’t include the total number of impressions so that information has to be copied from the Key Metrics tab and pasted into your working document. Feel free to paste over the ‘shares’ column since that doesn’t figure into the Web Liquid equation. Just make sure the impressions line up with the correct message.
4. Sort by content type and calculate sums.
Once the messages are sorted by content type it’s easy to add those selected cells to provide a sum of likes, comments and impressions for each type. It might help to break them into four separate groups just to keep the numbers straight per content type.
5. Calculate engagement.
It’s easy to calculate engagement the same way it was done in the study. Add together the likes, comments, and shares and divide by the number of impressions. The final figure roughly displays the percentage of people who took action — or were engaged — because of a post.
The study isn’t perfect but for those unsure of what content is working for them, it’s a great first step to understanding.
So how do these figures align with what you’re finding?
Related posts:
Thanks a lot for this post, Patrick!
Haven’t made my calculations yet, but these findings are well in line with the analysis of ‘One Week of Posts by 601 College Facebook Pages’ by Varsity Outreach in September 2010: http://blog.varsityoutreach.com/2010/09/analysis-of-one-week-of-posts-by-656.html.
They calculated engagement differently at the time, but the results they got were pretty similar:
Average Engagement Score
All Posts: 2.27
Posts with a photo attached: 3.62
Posts with a video attached: 2.99
Posts with no media attached: 2.35
Posts with a link attached: 1.67
Posts with music/sound file attached: 1.14
Another great post. The next big head-scratcher for marketers will be determining how the “amount of engagement” (as defined here) measures up to the “amount of awareness” (as defined in the traditional sense). Many of the measurement/operating models and methods for paid-media communications (advertising, etc.) are based on things like reach, frequency and impact. Although completely different in many ways, there is a need for similar “engagement equations” for social media to be developed so that strategists, planners and community managers can really understand the value of inciting engagement – and move beyond the premise of “more engagement = good.”
Thank you for this information. Kindly consider adding some images.
Inga, thanks for sharing the data and surprised it aligned so closely with what Web Liquid found.
Great points all around, Dave. If all the engagement centers around negative comments, does more engagement = good? I wouldn’t think so. It would also be nice to be able to value certain engagement in a way that it could relate to those traditional metrics for paid-media communications … but I imagine those values will all have to come, of course, from the goals and strategy in place.
Great information. It is important to make sure to respond to even the negative engagement and try to fix the issue. Hopefully it will prevent the same thing happening in the future and make your customer happy again.
Hello Patrick
I was reading your article and i wanted to know about calculating the engagement rate for the month of February for one of my Facebook pages When i download the post level excel sheet… you said “Add together the likes, comments, and shares and divide by the number of impressions. ”
However which likes and comments and shares are you referring to?
There are the ones in Lifetime Talking about this Sheet
there are the ones Lifetime post stories by action type
Should i consider both of them and add them all up ?
Thank you