“The most difficult thing in life is to know yourself.” –Thales
As purveyors of social media knowledge, we believe it is important to evaluate our own work, as we do others’.
To learn more about ourselves and our corner of the world, we mined all 6,241 tweets that @PRDaily sent in 2017 and analyzed the data. Many of our findings aligned with agreed-upon best practices. Some didn’t.
If you read no further, take this fact home with you: No two social media accounts are alike, and what works for others may not work for you.
Here’s what we learned from 2017:
1. Tweets mentioning well-known brands got the most clicks.
Some 40 percent of the top 25 tweets included a brand name (United Airlines, Starbucks, Google, Facebook, KFC, etc.), whereas only 21 percent of tweets overall mentioned a brand.
2. Infographics rule.
Tweets with the word “infographic” got roughly double the average number of clicks, retweets and “likes” compared with the average post.
3. Millennials are still “in.”
Any tweet using the term “millennials” yields about four times as many retweets as an average post, though clicks were roughly the same.
4. Statistics increased reach more than they drove drive traffic.
Nine out of 25 of our most retweeted posts of the year involved a percentage. That’s 36 percent, while just 3.4 percent (214 out of 6,241) of our tweets contained a percent symbol (%) or the word “percent.” These posts also had an average of 12.8 retweets, well above the average of 4.7.
Posts with a percentage were just barely above average in terms of clicks, however, averaging 20, compared with the overall average of 17—well within standard deviation.
What is the lesson here? We speculate that we gave away the crux of the story in the tweet, leaving no reason for the reader to click.
Every social media marketer should have a set of goals for each social network. To see actionable results, make sure you tie your business objectives to your social media goals.
If your goal is to drive traffic to the website, leave something to the imagination. If you want to spread your reach on social media, give readers the story up front.
5. The best time to post is different for everyone.
Here’s an unsatisfying insight: The best time to post depends on who you are. According to a Hootsuite analysis of 40,000 tweets, the best times to post to increase clicks and retweets is between 3 and 6 p.m., but Sprout Social recommends posting from noon to 3 p.m. of an unspecified time zone.
However, we found that across the board, the evening (8 p.m. to midnight Central time) was the best time to post to increase clicks, engagement and “likes.” By far, 10 a.m. to 1 p.m. Central time was the worst.
6. Data can lie.
Of the 25 tweets with the highest engagement, 19 were replies starting with a handle—meaning they were not visible to our general follower base.
Because engagement rate is measured as a function of “likes,” RTs and comments per impression, posts with minimal engagement got an outsize engagement score, because so few people saw them.
This was our tweet with the second-highest engagement of the year. (The first involved a Kardashian, which is a whole other ball of wax.)
Thanks! You’re the best.
— PR Daily (@PRDaily) April 20, 2017
If you’re not impressed, you should trust that instinct and always question data.
Just for fun, here were our superlatives of 2017:
Most clicked and most retweeted:
#United offers a primer in how NOT to handle a crisis.
— PR Daily (@PRDaily) April 11, 2017
Dear Mr. @Scaramucci,
Kindly take us up on our offer to attend one of our PR events for free.
— PR Daily (@PRDaily) July 28, 2017
Survey: Early bird gets the media coverage worm
— PR Daily (@PRDaily) July 18, 2017
You can replicate this experiment with your own Twitter account. Go to the native Twitter account, click on your profile pic in the upper-right corner, then Analytics → Tweets → export data. You can download about three months’ worth of data at a time. Then you can compile them into a master spreadsheet.
Note that the posting times for Twitter are gauged in Greenwich Mean Time by default, so anyone in the U.S. will have to adjust.