While there’s certainly a wealth of data to be gleaned on the Facebook Insights tab, a color-coded analysis on the Excel export can teach you a lot very quickly about what’s working, what isn’t and why.
What I love about this method is it’s actionable. Just a few minutes of sorting and color coding can help you optimize or entirely rethink your Facebook content strategy.
Here’s how to do it:
Export your Post Data (not Page Data)
On Facebook Insights, select your desired timeframe and export your analytics by Post Data, rather than the default Page Data, as you’ll be analyzing the metrics on each of your individual posts.
Skip to the “Lifetime Stories by action” tab (third tab in)
This will serve as your primary tab for the analysis. You may want to move this tab to a new workbook for easier access.
Copy and paste from “Key Metrics”
Skip back to the first tab, “Key Metrics,” and copy and paste the following columns into your primary workbook (or the “Lifetime Stories by action” tab).
- Lifetime Post Total Reach (the number of people who saw your post)
- Lifetime Post Total Impressions (the total number of times your post was seen)
- Lifetime Engaged Users (total number of people who clicked anywhere in your post)
- Lifetime Post Consumptions (total number of clicks on your post)
Make sure you do not do any sorting or rearranging before you finish the copy and paste portion of the analysis so as to not disassociate the metrics with their posts across tabs (all are automatically arranged by date).
Also, you’ll need to delete the second row before pasting from the “Key Metrics” tab, as this is a definition row that doesn’t exist on other tabs. The columns won’t line up until you get rid of it.
Copy and paste from “Lifetime Post Consumptions by type”
Copy and paste all of the metrics columns from this tab into your primary. Will include:
- Link clicks
- Other clicks
- Photo Views
Calculate engagement rate
Add a final column to calculate the engagement rate. Divide the total number of post consumptions by the total number of impressions.
Highlight the top 20% for each metric
Freeze the top row and then, one at a time, sort each metric in descending order and highlight the top 20%. So, if you have 50 posts, highlight the top 10. Pick your color strategically, as you’ll be color coding later. I recommend green.
Do the same for the bottom 20%
Repeat the process, but in descending order. Highlight in red.
Color code your top and bottom posts
You should have highlighted the top and bottom posts for approximately 10 different metrics. Choose a different color, say purple, and highlight the posts that are green in the most columns. These are your top posts.
Repeat the process for your bottom posts. Let’s make these blue.
Now you should have something that looks like this.
First, take a moment to read through your purple (top) and blue (bottom) posts with fresh eyes. What groups these posts together, and what sets them apart? Pay particular attention to:
- Post type (image, link, video, status update)
- Length of post
- Time and day of post
- Voice and tone
Do all or the majority of your top or bottom posts have any or most of these characteristics in common with each other? Look for links anywhere you can, but also pay attention to what sets them apart from each other. Could you have implemented a tactic you used on one post to make another one stronger?
Try to form theories as to what made each post successful, understanding that your insights will remain theories until you test these tactics in future posts.
Make sure to account for any advertising you did in your analysis. These posts will likely make it among your top performing because of sheer volume. Either disregard completely, or pay attention to the engagement rate to normalize across non-promoted posts.
Support your theory with green and red posts
Your green and red posts represent posts that performed very strongly or very poorly when measured against one metric, but not all. This means there is something making them successful (or unsuccessful), but something else keeping them from scoring in the very top or very bottom.
You can learn a lot from these discrepancies. Look at the same elements of these posts as you did above. Compare them to each other and to your best and worst performing posts.
For example, if all of your top performing posts were published on Tuesday, and your green posts were all published on different days of the week, you can reasonably conclude that publishing them on Tuesday may have given them a boost.
But something else besides the date allowed your green posts to outperform others in some categories. Look at the metrics that are green as opposed to not highlighted or other colors. Which metrics performed well? Were they engagement metrics or reach metrics? What about that post sticks out to you as similar to other high performing posts, and what seems different?
Here are some common discrepancies to look out for:
When a post with a top engagement rate doesn’t rank in the top performing posts across the board
Because engagement rate is the number of engagements divided by the number of impressions, this likely occurs for one of two reasons. Either the post had a very low reach, or the post was highly engaging in some aspects but not others.
The former is probably a visibility issue. The content was likely top notch, but people just didn’t see it. Pay attention to the day of the week and the time it was published. This is likely a low volume time for you.
Or, look at the date it was published. Was something major going on in the news that would’ve taken up real estate and crowded your audiences’ timelines?
If reach and impressions are not the issue, that means your post garnered a high number of engagements in one column but not another, which brings us to our next point.
Your posts performed strongly either in terms of clicks or engagements (likes, comments and shares) but not both
There are different psychological triggers that inspire a person to click on a post verses engage with it. The best posts motivate the audience to do both, though there are times when you might be striving for one type of engagement but not another.
People click on posts that pique their interest. Either the media or your text convinced them that this is something they want to learn more about, whether in the form of reading an article, watching a video or clicking through a photo album.
People engage with posts that impact them emotionally. They like, comment and share because they feel some sort of emotional connection, whether it be a humanitarian issue they hold close to their heart, a representation of a lifestyle they strive to be a part of or just something that makes them laugh.
If your post performed significantly stronger in one category than the other, assess if it’s for one of the above reasons. Think about how you could incorporate both curiosity and emotion into your content rather than just one or the other.
Disproportionate number of likes, comments and shares
Not all engagements are created equally either. I was surprised when I first noticed this, but I’ve seen it across the board for a number of clients (maybe you’ve had a different experience?). Often times, the posts that have the highest number of engagements in one category have a mediocre or low number in another.
While some people feel compelled to like, share and comment on a post all at once, this is often not the case. Likes will almost always outnumber comments and shares (it involves the smallest commitment on behalf of the user), but comments and shares can creep up quickly in the appropriate environment.
Try to figure out why that post ranked in one category but not all. Was there a dynamic conversation going on that attracted multiple community members? Was the content particularly share-worthy? If you can pinpoint why one category took off, you can start to understand why your audience behaves how it does, and adjust your social strategy accordingly.
A high engagement rate with a low reach
Your engagement rate is just your total number of engagements divided by your total reach – so if you have posts with unusually low reach or impressions, each engagement counts significantly more than a post with a great expansion.
However, this can also be an indication that your post reached a very niche audience that responded positively. For example, I find this is often the case for one of my clients who, though they have a national presence, often promotes events in their hometown. The posts don’t travel far, as they only pertain to a fraction of our audience, but among that fraction they perform very strongly.
This can also be the case when you tag someone in your post. While these posts have the potential to reach a large audience, it doesn’t mean they necessarily will. But they are magnets for engagement among your joint audience.
What have you found when analyzing your Facebook Insights? Do you find the data on the page to be more useful, or do you prefer to export and work with raw data like I do? Tell us in the comments below!