We analyzed 100m posts so you don’t have to
Facebook analysis is critical in a world of ever-changing social algorithms. In 2016, Facebook announced a series of changes to its newsfeed algorithms. The tweaks were designed to emphasize content from friends and family, reduce clickbait, and deliver content each user would find informative.
They also sent a clear signal to content marketers: the path to organic reach is narrower than ever, and it depends on serving your audience content they will find engaging. Two of our favorite experts summarized this well:
“Study what content generates most shares and engagement (both on your own page + others in your niche), and focus on publishing that type of content.” Mari Smith
“More engagement means more visibility in Facebook’s news feed. Facebook’s algorithm is more likely to give more visibility to posts that resonate well, to audition it in front of more people.” Larry Kim
Clearly, marketers need data about the types of posts that their audience engages with. But, beyond that data, they also need a clear process for Facebook analysis to discover what the data means.
Canva provides a great illustration of the analytical process in this post describing how they used data from BuzzSumo to discern the essential ingredients in successful image posts at BuzzFeed.
We thought it would be helpful to go one step further and lay out the basics of Facebook Analysis, so that you can more easily find, analyze and apply data to your posts.
Now, you can find the answers to the fundamental Facebook marketing question:
Why did this post get so many shares, while this one got so few?
Actionable insight is the key to unlocking your organic (and paid) reach potential on Facebook. Here’s how to get it with BuzzSumo’s Facebook Analyzer.
There are three categories of posts you should scrutinize: posts that relate to your industry and keywords; your competitors’ posts; and your own posts.
For example, these are the most engaging posts this year for the topic “Exams” on Facebook.
I’ve added a filter which limits my results to only those posts in English.
BuzzSumo defines engagement based on the sum total of interactions (shares, likes, and comments), but there is also an option to filter by the type of engagement you are most interested in.
Step one: done.
The next step would probably be more accurately described as analysis, but I find it more helpful to approach the task as a set of questions. What do I see? How are the posts alike? How are they different?
In this phase, I want to discover what, if anything, these successful posts have in common.
It’s tempting to just skim the results. But, slowing down to notice creates big dividends.
For example, you can see that all of the most engaging posts about exams are tapping into common shared experiences: procrastination, pressure to perform, fear of failure.
The images are memes, and there are many pop culture references.
However, not all searches will be this easy to analyze, or this fruitful.
For example, here are the results for a search for Halloween:
I looked carefully and found that posts about costumes, party foods, and holiday decor got the most engagement. That’s pretty much what you would expect for any holiday, though. So the information is not that helpful.
To find more useful results, narrow these types of searches based on the big picture categories. I.e. with, Halloween, I can change my search parameters to “Halloween costumes” or “Halloween decorations.”
Here are the results for my search for “Halloween decor.”
Now, I can quickly identify two themes: DIY decorations, and images that inspire by giving a room wide “look” or visual vignette. Much more useful insight.
Once you’ve done your own quick evaluation of the successful posts, click on “Analyze.”
At a glance, BuzzSumo will show you the most popular type of post, the most popular post length in characters, and the day and time when people interacted most with posts about the topic.
For the “Exams” search, here are the results:
The most popular length is 50 characters, just one-third the length of a tweet.
The day and time when these posts got the most engagement is also interesting — Sunday at 4 a.m. UTC time, around midnight my local time.
I picture people up late studying, trying to focus on the night before a big exam. They log on to Facebook for a little solace. What do they find engaging — posts that tap into the theme of procrastination, posts that remind them they aren’t alone in suffering through a tough exam session. Posts that reduce stress by making them laugh. This result makes sense, but I’m a bit puzzled by the “Giveaway” as the most successful type of post.
So, I go back to the tool and filter to find the most successful Giveaway posts.
In this case, it’s Cisco Learning leading the way:
That insight is helpful if I am targeting adult learners, or people who want to be certified in a new skill.
All of this is just an exercise in wasting time if I don’t draw some conclusions I can use.
I need to step back and ask:
What did I learn?
What will I try?
In other words, I need to apply the insights and test to see if they hold true for my audience.
If I want to reach adult learners interested in exams, I can create a giveaway offer and experiment with promoting it via Facebook.
If the Cisco audience isn’t my target, I can circle back, filter out Giveaway results and look at examples of the next most popular post types.
Data-driven insights tell us where to start. Continuous evaluation tells us how to proceed.
BuzzSumo helps makes evaluation easier.
The data complements the insights available from Facebook for pages you manage, and the scope of the information enables you to get a big picture look at what your audience prefers.
For example, when I search for the most engaging posts at the BuzzSumo Facebook page, it’s clear that our own blog posts get the most attention:
But, I also notice that some posts of others’ content show up in the results.
So, I try to learn more, sorting by different types of interactions. When I arrange the results based on shares alone, a post where we featured a customer was third; a controversial research post first.
The takeaways: [Tweet “Don’t be afraid to be controversial or to give away coveted Facebook real estate.”] The five shares from the post that featured our customers were a form of earned media as our owned content, the original post, was leveraged via shares into others’ news feeds.
I can adjust my content and social strategy accordingly, adding posts that call attention to our customers’ brands and successes.
Another great way to use the page search feature in BuzzSumo is to evaluate what is working for your competitors’ or industry leading pages.
Unlike Facebook, BuzzSumo makes the same insights you can access for your own page available for your competitors’ pages.
Knowing what works for them, analyzing that information to tease out actionable insights, and applying the principles to see if your audience will respond is a best practice for competitor intelligence.
Your research can be simple.
Searching, asking, and making note of the analysis provided by BuzzSumo.
Or it can be more complex, drawing comparisons between several competing pages and your own page.
For example: This spreadsheet compares a new beauty industry passion page to other established beauty companies’ Facebook pages.
I also used BuzzSumo to determine the post categories that got the most interactions and evaluated each of them, asking myself what message they contained, what their focus was, and if there was a call to action.
Categorizing the results helped me to see patterns and differences in emphasis for the pages.
I used the BuzzSumo Analysis tool to find and compare interaction levels with the pages for different time periods.
I also compared the best day and time for interactions with each company’s posts.
At the end of the process, I was able to draw conclusions about each of the pages I studied.
For example, here’s the summary for the Mary Kay Facebook page:
The interaction data confirmed that this is the strongest page of those I evaluated, and the summary gave me great insights I can use in developing my own posting strategy.
Because I was able to gather the same types of summaries for the other company pages, I can also target posts from the passion page to lookalike audiences that fit the style of my posts.
None of this analysis was overly complicated. It just took an intentional effort to find relevant data, ask myself what the similarities and differences were, and make some strategy decisions based on those insights.
The final step in the process is even more simple.
If a particular insight drawn from your Facebook analysis doesn’t generate results for your audience, change it up. To be fair, it’s worth testing each type of post more than once, but if you consistently get poor results, don’t be afraid to change your strategy.
Your own experiences form the next round of data available to you, and you can begin the Facebook analysis process again:
Enter any topic, term or url to search to see BuzzSumo in action. It’s free!
100% free. No credit card required.