Blog post written by Anders Mikkelsen, Analyst at Mindjumpers
Facebook Insights is a treasure box of information for community managers who want to see the effect of their brands’ wall posts, and optimize on future content. But where should you start in the masses of available data about your posts? In this blog post I will look into the post level data that community managers can export to Excel sheets via Facebook Insights. I will share a few tips as to which questions and answers to use when analysing the data and show how to use your new found findings to optimize on future community management.
Step 1 – Show me the data
In Facebook Insights you can export an Excel post level data sheet with 7 tabs and almost 50 types of data on all the posts you have published in a given period. So, which part of the data is relevant to you then? Well, of course this depends on your brand, your industry and the content you post, but overall there are data parts that are important to every brand who posts content on their Facebook wall.
Generally, the data we look for when viewing and analysing for our customer brands in Mindjumpers are the following:
- Reach (number of unique people who have seen your posts)
- Engaged users (number of users who have engaged with your posts)
- Photo views
- Other clicks
- Link clicks
- Video plays
- Unlikes (note: this data is not in the post level reports, but in Facebook page level reports)
- Talking About This
- Engagement rate (engaged users/reach)
- Virality (talking about this/reach)
Using this data a typical monthly scorecard for a brand could look something like this:
Step 2 – Let the analysing begin
So far so good. Now you have a scorecard. But measurement in itself only has a value if used constructively. So now it’s time for you to analyse on why the numbers look the way they do. Here are some hands-on examples of questions and answers to ask yourself when analysing:
Q – Why does one of our photo posts have an extremely high number of engaged users?
A – The particular objective on this photo is apparently more interesting than other photos.
Q – Why is there a large number of unlikes on the publishing date of a specific post?
A – Maybe there is something in the post that irritates the users.
Q – Why are there no plays on one of our video posts?
A – Maybe the video content is simply quite boring.
Q – Why do some of my posts have a very high reach and some not when the content is similar?
A – Maybe it has something to do with the posting time.
These are some examples of how to think when analysing the data. The most important factor to look at when searching for your answers is the quality of the content. What type of content is it? Is it exciting? Does it call to action? Does it have a news value? Does it make users want to engage with each other? Another very important factor is the date and time of publishing. It is not unlikely that there is a time pattern between the posts that perform well and the posts that don’t.
Step 3 – Optimise on future community management
You now have a good impression of which posts work and which don’t. It is then time to put your new knowledge into action. This is pretty simple. Look at what post types (updates, photos, videos, links, questions) perform the best. Look at what text and media that gets the most engagement. See if there is a publishing time pattern between your most successful posts. Learn what works from your best posts and make more of these good posts. Meanwhile, minimise the type of posts that have previously performed the worst for your brand. Use these insights in your future community management, and get ready to see your posts perform increasingly better.