March 01, 2017
TWITTER TESTS The best way to understand your Twitter audience and build better content around your brand is to test appropriately. This will help you move forward, and make decisions that turn dayto-day head-scratching scenarios – When should I post this? Should I be posting right now? When does this post cease to be relevant? – into total no-brainers. Read on to learn about our three favorite tests.
22. THIS MANY TWEETS VS. THAT MANY TWEETS Why: Your cadence can have a big impact on your campaign performance. Too few Tweets can make you seem uncommitted and too many can feel like spam. So, how many Tweets is just right? Try this: Tweet for one week in 15-minute increments and another week in 30-minute increments. After running this test, examine all the data from each week. You can use the Simply Measured Account Report or Google Analytics to get this data, export it into Excel, and compare the weeks side by side. Make sure your sample set excludes variables like business hours, holidays, and other times your brand does not tweet. For example, if you only tweet in 15 or 30-minute intervals between 6 a.m. and 5 p.m. PST, but you’re reporting on a full 24 hours, you may find that your data is skewed after the end of the business day.
24 Ways to Test Your Instagram, Facebook, and Twitter Content 28
23. THIS TIME VS. THAT TIME Tweeting in 15-minute intervals, everyday, between the hours of 5 a.m. and 9 p.m. would produce 64 Tweets per day. Producing such a large number of Tweets can require a lot of effort and may or may not work for driving traffic to your website. That’s why it’s important to prioritize times and days to Tweet rather than trying to cover each hour of the day evenly. The key metric to test for site traffic is click-through data from Twitter to your brand website. Try this: Identify how many Tweets were sent each hour. This can be done manually or by using a Simply Measured Twitter account report, but the end result you’re aiming for is a bucket of Tweets for each hour of the day. When gathering your data, make sure your sample size is significant enough that it can capture the full scope of your efforts on Twitter. From there, bucket your website traffic from Twitter based on those hourly segments, and come out the other side completely insightful. How can YOU do this analysis? Check out our Building A Better Twitter Calendar guide here for a full step-by-step breakdown.
24 Ways to Test Your Instagram, Facebook, and Twitter Content 29
2. THE RATE OF DECAY Why: The words “engagement rate of decay” may sound frightening, but it’s really fertile testing ground. How many times can a piece of content be sent to the same audience without fatiguing them (and, even worse, making them click the Unfollow button)? Try this: Patrick Ruffini, the founder of Engage, a digital agency in Washington, DC, conducted an experiment where he tweeted the same call to action 10 weekdays in a row, at the exact same time. How did this affect engagement? Did engagement die down after the first Tweet, or did engagement increase as exposure increased? Here’s the Tweet that Ruffini shared, every weekday for 10 days at 3:36 PM:
What did he find? He charted signups to the lead gen card here:
Ruffini found that submissions dropped off significantly after the first Tweet, but didn’t die out completely. The first Tweet generated a 72% conversion, meaning the number of people who saw the Tweet and took action. The average across all 10 Tweets was 58%. As Ruffini put it, “Even if you limited the repetition to five Tweets, 65% of the exposure would have come from Tweets 2-5.” Ruffini’s experiment provides a good example for the types of tests a brand marketer or community manager could use to test Tweet repetition.