Dan Zarella recently released a sweet, scientific infographic on Facebook statistics titled How to Get More Likes, Comments and Shares on Facebook (UPDATE: that link is now broken, please read more on Llikes here). I am big fan of Zarella’s as he is dedicated to the data and science and the actual behaviours of people with social media and the web. But data and science by themselves do not necessarily help you. They are are just, well, data and science. It is the application of this information that is useful and it is only useful to YOU if it works for what YOU are trying to accomplish with YOUR audience. Here’s a potential process you can go through to apply rich data like this to improve your strategy.

1. Read It! (I mean really read it)

Sounds simple I know but the tendency with processing data is to gloss over it. Infographs help distill the key facts but also make it easy to scan, think “that’s cool” and move on. Go over it a few times, see if there’s an explanation of the process or other relevant information on how the data was accumulated.

2. Recognize It!

There are three things to “recognize” as you look over the data.

  1. Things that you should be wary of. For example, in the infographic below, the data is from the 10,000 most “Liked” pages many of which are for-profit organizations. Since my main focus is on charities I’ll approach this data a bit more cautiously as it is generated from outside of my focus industry. Another example of this would be equating “sales” and “fundraising”. Great amounts of cross-over and much we can learn about fundraising from sales but they are still different. With things that you should be wary of… approach with a caution.
  2. Things that reinforce what you would assumeFor example, negative posts generated more comments. That makes sense. If a post is negative it is more likely to “stir the pot” and evoke an emotional response leading to a comment. When data reinforces what you would assume that is a good thing as this is the most reliable type of data. Another example may be seeing a study about Millennials and giving and how they are heavily influenced by their peers. Logical, makes sense with what I know and backed up by data. With things that reinforce what you would assume… approach with confidence.
  3. Things that contradict what you would assumeFor example, photo posts having the greatest amount of likes, shares and comments. My assumption would’ve been videos. After giving it some more thought this actually made sense as videos take extra time to watch, something many people are not willing to do so photos provide the visual stimulation of a video without the time commitment. Another example is the best times to post in terms of day and time of day. I would’ve assumed during the week around noon would’ve been better but after thinking about it this makes sense to me as well. People are busy during the day and generally have a bunch of things going on so they are not on Facebook or not fully present on it and in “surf” mode. Cruising through Facebook on the weekends or after work as a way to unwind and distract from work seems logical and why engagement metrics might be higher during those times. With things that contradict what you would assume… approach with creativity.

3. Formulate It!

Alright now you’ve read through the data and recognized some key areas. The next thing you need to do is determine what makes sense for you and what you are trying to accomplish. Generally, the things that contradict what you would assume are where you can see the greatest gains with a slight change in strategy and approach and why you should approach those with creativity. I’m focused on helping charities build an audience and engaged following so more likes and shares on Facebook is very important. When I see the data on how weekend and evening activity is boosted I’m very eager to see if that works for the organizations I’m working with.

4. Test It!

So from the data, I’m curious if the audience I’m working with will respond the same way to evening and weekend posts so I need to figure out a way to test this. In this case I would try to get a few weeks worth of data on “normal” posting schedules (weekday and day times) with regular content types (photo, video, text, blog link) and then get a few weeks of data on the “trial” posting schedules (weekends and evenings) with the same content types (photo, video, text, blog link). The more variables you can keep constant the better. Try to use similar titles for blog posts (use a question as the title for each for example) and content types (photo of a client) so when/if you see differences in response it is more reasonable to make the connection between the timing as the driver behind the changed behaviour and not something else.

5. Use It!

Were you satisfied with the test results? Was it hard to tell? Running the test again or doing iterative testing (where you slightly change one variable from the past test) might make sense but once you feel comfortable that your hypothesis has been proven true or false go ahead and start implementing the new tactic as part of your strategy.

Finally, you should always keep looking for new data as the world (and web) is always changing so when new data comes out in 6 months go through this same process so you can let your data guide and drive strategy, not throw it into chaos.