The best things we learned from our latest Twitter experiments

Regular readers of our blog and viewers of our office hours know that we love to encourage a healthy amount of experimentation in your social strategy across platforms. It’s wise to always work from best practices and with the knowledge your data is giving you, but also leave room for ideas that grow out of new trends in your industry or that might come from anywhere else.

Experiments can be spontaneous, or they can be very well-planned out. Often the latter gets a little bit of the former thrown in, and that’s great too.

In fact, that’s what happened with the experiment we’ve been running on Twitter since the summer. We thought we’d share some of our takeaways with you here.

Got questions or something we missed? Find us on Twitter @UnionMetrics.

Decide on your experimental parameters

While it’s impossible to isolate every variable in social media, you want to do things as cleanly as possible so you know what’s affecting the measurements that are important to you and your brand (we’ll cover that in a minute). In our case, we laid out a series of variables we wanted to test, one at a time, over two week periods. We decided on all of them ahead of time, then during the course of the experiments, thought of new ones we wanted to test and worked them into the schedule.

We also decided which metrics we would measure at the end of each two-week period to compare and see which variables were increasing our impressions and engagement, and which were decreasing one, the other, or both.

We used our comprehensive Twitter analytics to do this; our Tracker has been gathering data around our Twitter account for years and it’s easy to narrow down each window in time to get the exact metrics we need for the exact dates we needed them.

Decide what you’ll do with the results

Chances are if you’ve been running social media for your brand for a while you have a sort of routine that you stick to. An experiment like this is the perfect way to see which parts of that routine are still serving you and which could stand to be refreshed.

For example, we found that posting more often had a negative impact on our impressions and our engagement. We set that experiment up to tweet more during hours we normally did not tweet, however, so that tells us that we don’t have fans, followers or customers awake and wanting to engage during those hours. (It’s always worth checking in a global market.)

In the future we could try tweeting more often during our normal time windows to see what kind of impact that has, or even running this same experiment again in a year or two. Audiences are constantly shifting which is why it’s important to constantly check in and be sure you’re meeting their needs.

Get creative

Obviously use the data you get to tweak your everyday routine, but also use it to inform the content strategy and calendar you’re planning for 2018. Take some time to sit down with your team and brainstorm around the results; you might get some exciting new ideas out of it to bring into next year.

Combine those with the latest trends in your industry and platform features to really maximize the potential of your brand’s presence across platforms.

Want a demo of our analytics?

If this experiment sounds like something your brand could really use as 2017 winds down and 2018 nears on the horizon, we’d be happy to give you a personalized demo to show you exactly how our metrics can work for you. Get started below.

Specific, actionable social intelligence. See how Union Metrics helps marketers win.

Request a demo