Facebook is the greatest platform animal advocates have for exposing the cruelty of factory farming and promoting vegan eating. The more people we reach with our posts, the more animals we can help.
But how do we do this? How can we reach more people with each post?
To find out, we analyzed data from 1,600 Facebook posts created by farmed animal advocacy groups. All told, these posts were viewed over half a billion times. Yeah, that’s right: half a billion.
We looked at every share, every like, every comment, and a heck of a lot more.
We then crunched the numbers and found out what makes a farmed animal advocacy post go viral. We also found out what makes a post flop. And now we’re going to tell you.
No time to read now? You can still download and print our Nine Rules for Dominating Facebook for Farmed Animals and tape it by your desk as a daily reminder.
Thanks to Vegan Outreach, Compassion Over Killing, and The Humane Society of the United States’ Farm Animal Protection Department for sharing six months’ worth of Facebook data, which was used in addition to Mercy For Animals’ own data. Thanks to Krystal Caldwell and Sarah Severson for their data collection and analysis, and to the funders that made this study possible.
What is a Successful Post?
We wanted to learn which posts perform best on Facebook. You could define best performance in a lot of ways: the number of comments, likes, shares, impressions, clicks, and so on. We measured all of those, and you can find all the data in the “Numbers” section at the end of the report.
But what seems most important to us are impressions. Impressions are the times someone sees your awesome post. The more people see it, the more people you can persuade to change.
We also think impressions by non-followers are particularly important. After all, we don’t want to just preach to the choir. We want to reach the meat-eating masses.
Now, how well a post performs on Facebook isn’t the only thing that matters. How impactful a post is in making someone eat vegan (or whatever) matters too. A powerful post that reaches a thousand people can do more good than a weak post that reaches two thousand.
Here’s how we see it: Impressions x Impact = How Good a Facebook Post Is
This study looked at which posts got the most impressions. It didn’t tackle impact, though we hope to do that soon.
So keep impact in mind as well when deciding what to post.
Okay, we’re done babbling. Here’s how to dominate Facebook and save millions of farmed animals.
First, we’re going to give you our Nine Rules for Dominating Facebook for Farmed Animals. This has everything you need to know in a nutshell. Save it. Print it. Tape it to your desk. Refer to it often. If you do, animals will love you.
After that we’ll give you the details on what our study found. Prepare to be absolutely amazed at how well the Rules work.
Nine Rules for Dominating Facebook for Farmed Animals
- Post tons and tons of Facebook videos. Then post some more.
- Make people feel something, especially sadness, anger, amusement, or nausea.
- Use fewer than 10 words.
- Ask for shares, not likes or comments.
- Quote often.
- Link to a news article, not a blog, when you want more traffic.
- If it’s about an animal, make it a baby.
- Farmed animal + companion animal = win.
- Post food and meme photos.
Hey, you: you’re going to forget all this!
So please download the PDF of these Nine Rules for Dominating Facebook for Farmed Animals, print it, and tape it by your desk as a daily reminder.
Type of Post
Our results matched the findings of general Facebook studies.
The exalted masters of Facebook are Facebook videos. They generated 130 percent more impressions than photos (and 170 percent more impressions among non-followers).
The data used for this study is from 2014 when Facebook was tweaking its algorithm to favor videos. But it appears Facebook is still favoring videos to this day in an attempt to make YouTube irrelevant.
So pretty please: post a lot, lot, lot more Facebook videos!
Beyond that, photos beat links. Photo posts of any kind got 20 percent more impressions than links (links to blogs, YouTube, or anything else). Percentages were twice as big when it came to impressions by non-followers. Their only downside was that they had more un-follows than other types of posts.
An Important Reminder...
You know how we said photos beat links? Well, we just mean they got more impressions. That doesn’t mean you should stop posting links.
Like we said earlier, how many impressions a post gets isn’t the only thing that matters. How powerful the post is in making someone eat vegan (or whatever) matters a lot too. An inspiring link could change more diets than a cute photo, even if fewer people see it.
Raw Data and Analysis Tables
Remember: Impressions x Impact = How Good a Facebook Post Is
Got it? Good. Let’s keep going.
Emotional posts did much, much better than neutral ones.
Sad, nauseating, funny, and angering posts got 80 percent more impressions than neutral ones. Among non-followers, they generated 150 to 200 percent more impressions. Awesome and cute posts also did fairly well. So make people feel something!
One word of caution: Sad and awesome posts also had a higher percentage of un-follows. This isn’t too big a deal though; un-follows are still rare.
Another word of caution: There’s anecdotal evidence that Facebook’s algorithm is now punishing unpleasant content and favoring happy content. Sad, nauseating, and angering posts may not work as well today as they did in the past.
Use fewer than 10 words in your Facebook post whenever you can. When it comes to text, short is sweet.
For example, posts with 10 words or fewer generated 30 percent more impressions than posts with 11 to 25 words. The percentage was even higher for non-followers. And the longer the text, the worse things got.
(We’re talking about the text typed into Facebook. Words in a photo meme or automatically displayed for a link don’t count.)
Call to Action
Telling people to share a post generated 20 percent more impressions than not including any call to action (45 percent more among non-followers). Plain old “share” and “please share” worked fine. “Read and share,” or for petitions, “sign and share,” may have worked even better.
Don’t tell people to like your post or comment. In our study, posts that encouraged likes or comments had fewer impressions than posts with no call to action.
It’s also pretty clear that telling people to “donate” slashes impressions. But if you do social media work for a nonprofit, you already knew that.
Quick note: Most of our data on call to action does not reach statistical significance. But it’s likely accurate. And general Facebook studies have gotten the same results.
News flash: the Internet loves happy animals!
Facebook posts linking to blog posts about happy animals, cruelty news, and food product news all generated more impressions than Facebook posts linking to blog posts about general vegetarian news or other topics.
But when it came to driving clicks, blog posts about happy animals ruled. They got 165 percent more clicks on Facebook than blog posts about general vegetarian news.
Blog posts about food product news and cruelty news got 45 to 55 percent more clicks on Facebook than blog posts about general vegetarian news.
Quote people a lot. It’s easy and makes a big difference.
Posts with quotes got 40 percent more impressions overall and 80 percent more among non-followers.
Type of Link
Links to news articles got the most clicks and impressions.
They resulted in 60 percent more impressions than links to blog posts (130 percent more among non-followers). They also generated 80 percent more clicks.
Links to YouTube videos did not fare any better than blog links.
So if you want more people to read a story, link to a news article on it instead of a blog. Of course, you may have other goals. Maybe you want to drive people to your site so they sign up for your e-newsletter. Maybe you want people to read your version of the story because it’s more persuasive than CNN’s. That’s all well and good. Just keep in mind you’ll get fewer readers.
Type of Animal
Posts about pigs and cows got 25 percent more impressions than ones about chickens (45 percent more impressions among non-followers). Lamb posts may have done well, but we don’t have enough data to be sure.
There were no statistically significant differences between chicken posts and posts about ducks, turkeys, or fish.
Pair up a farmed animal with a dog or cat if you really want to go viral. These posts got 55 percent more impressions than posts about birds or fish and 100 percent more impressions among non-followers.
Age of Animal
Babies got more views. Posts with newborn or juvenile farmed animals got 20 percent more impressions than posts with only adults.
It’s fine to keep mom or dad in there too. Just don’t forget the baby!
Number of Animals
Posts about a large group of animals generated 20 percent more impressions than posts featuring an individual animal. We weren’t expecting that!
When it comes to photo posts, photos of food did better than photos of animals. Food photos generated 40 percent more impressions than animal photos (75 percent more impressions among non-followers). Photos of memes with text also seemed to do better than animal photos.
So you may want to start dishing more food pics and more sassy memes.
Time of Day
General Facebook studies have found that posts made at certain times of the day get more impressions. Supposedly 4 to 7 p.m. EST and 8 a.m. EST work best.
In our study, time of day didn’t matter. Maybe farmed animal advocates are different from the general public. We don’t know. Consider the jury out on whether time of day matters.
Day of the week also didn’t seem to matter when it came to how many impressions a post generated. Sunday posts seemed to get slightly fewer impressions, but the difference wasn’t statistically significant.
General Facebook studies have found using exclamation points increases impressions. In our study though, posts with exclamation points had 15 percent fewer impressions (and 25 percent fewer impressions among non-followers). So we don’t know what to tell you here.
Posts with question marks may have gotten slightly more impressions than statements.
Hashtags and Emoticons
General Facebook studies have found posts with hashtags or emoticons get more impressions. In our study, we didn’t see a statistically significant difference between posts that used them and posts that didn’t. If anything, it seems that posts with hashtags and emoticons had fewer impressions.
So consider the jury out on this question too. #uncertain
Specific vs General
We checked whether posts that focused on a specific incident (like Tyson abusing chickens) seemed to get more impressions than general posts (like “chickens lead miserable lives on factory farms”). We didn’t find a statistically significant difference between the two.
We found the same thing when comparing links about current events with links about general issues: no difference in the number of impressions.
If you’re not a statistician, analyst, researcher, or someone who wants to look at the data much more closely, please turn back now. Otherwise your head might explode from too many numbers and too much statistics lingo. Don’t say we didn’t warn you!
Okay, if you’re still here, it means you’re probably wondering how we did this study. You may want to make sure our data is sound. Or maybe you want to see all the data for yourself, not just the stuff on impressions. Good for you.
Here’s what you’re looking for:
MFA, Compassion Over Killing, Humane Society of the US Farm Animal Protection division, and Vegan Outreach exported data directly from their Facebook pages to provide most of the raw numbers. Six months’ worth of data (July to December 2014) was collected from COK, HSUS-FAP, and VO, and one year’s worth of data (2014) was used from Mercy For Animals. Complete data for nearly 2,000 posts was exported and compiled by MFA’s researchers.
The researchers then viewed and manually coded each of the posts across a number of different categories (like time of day posted, emotional tone of post, number of animals in post, and so on). Other researchers cross-checked the coding to ensure accuracy and uniformity. Posts that were geotargeted towards certain audiences or that were not in English were eliminated prior to analysis. A total of 1,655 posts were analyzed.
For each for the 8 outcomes, MFA’s researchers ran four log-linear regressions:
- Bivariate regressions for each of the outcomes on each of the 21 independent variables. Table 1 compares the results from each bivariate analysis to the results from the multiple regressions (see below).
- Multiple regressions for each outcome on 17 variables (animal type, animal number, animal age, hour posted, day posted, post type, amount of text, emotion, statement type, hashtags, emoticons, call to action, photo subject, quotation, specificity, organization, and month). The results from these specifications were reported in the text above and are available in Table 1 and 2. Four variables were eliminated from these analyses (link focus, link photo, specificity of link, and blog focus) due to the lower number of completed observations that were available for them.
- Multiple regressions for each outcome on 20 variables (the 17 listed above + link focus, link photo, and specificity of link). The results for the analyses for these four variables are available in Table 3.
- Multiple regressions for each outcome on all 21 variables. There were only 650 completed cases for all 21 variables (vs 1500 with 17 variables), so these analyses were only used to report on the relationship between blog focus and each outcome (Table 4).
- Tables 1-4 display regression coefficients as percentage points. Each coefficient is relative to a baseline, and is presented alongside 95% confidence intervals and its p-value.
Raw Data and Analysis Tables
Download the raw data here and the analysis tables here. If you’re going to look at the tables, read this first.
Take a look at the second sheet in the Tables and the rows near the top labeled “Animal Type.” (The second sheet is the main set of numbers you’ll want to look at.) As noted, the “Animal Type” baseline is “Chicken.” That means that posts about other animals are shown in relation to posts about chickens. The table is showing whether posts about other species did better or worse than posts about chickens.
For the “Animal Type—Cows” row, the “Impressions” column shows the number 24.1. That means cow posts got 24 percent more impressions on average than chicken posts.
If the number had been negative, it would have meant cow posts got fewer impressions than chicken posts. For example, look at the row for “Animal Type—No Animal Featured.” In the “Impressions” column it shows the number -13.99. Because this is a negative number, it means that posts with no animal featured got 14 percent fewer impressions on average than the baseline, which is posts about chickens.
Is it all making sense now?
Cells that are bold are ones that we’re very confident about. If a cell is a bold positive number, it means that it’s extremely likely that the row in question performs better than the baseline. If a cell is a bold negative number, it means that it’s extremely likely that the row in question performs worse than the baseline.
If a cell is not bold, we’re not entirely sure that the numbers are accurate. This is because either the results are very close to the baseline, or the sample size was too small.
On sheets 3, 5, and 7, you’ll see that some cells are green and some are red. We’re just using color here to represent the same likelihood as the boldface. Green cells mean it’s extremely likely that the row in question performs better than the baseline. Red cells mean it’s extremely likely that the row in question performs worse than the baseline.