7 things we learned from broadcasting full episodes of our PBS show on Facebook

The content you’re seeing on Facebook has changed a lot in a year. Back in October of 2015, we were seeing a ton of video content in our News Feed, and noticing the impact and reach of short, front-loaded, text-heavy video. In December, while we were developing our new debate show Point Taken, Facebook introduced Live streaming, which immediately opened up exciting new possibilities for direct engagement with Facebook’s 1.7 billion users in a way that was central to our mission of bringing civil dialogue to a generation that only knows talking heads and rigid polarization.

As Facebook Live matured, we saw broadcasts getting longer and longer, as it allowed more people to join the conversation live. And we wondered, could there be an appetite for our entire 26-minute show on Facebook?

Our episode debating “Should salaries be transparent?” on the left was uploaded as a native Facebook video; the episode “Is technology making us smarter or dumber?” (right) ran as a Facebook Live.

With the blessing of PBS, we set out to test that theory. We mixed up the 10-episode run by posting the full episode three different ways to Facebook — as a native video posted synchronous with the 11pm ET broadcast, as a native video posted early the next morning, and as an 11pm ET Facebook Live stream — all with different results. Here is what we learned:


A snapshot of the comment thread from our technology debate on Facebook
  1. Even though Facebook Live videos are enjoying an automatic “reach” boost, the type of engagement we’re seeing on Live videos are a game changer. The types of comments and conversations that people had on our Live streams were overall more substantive, thoughtful, and responsive to each other than on our natively posted video. We think this is because people understood the community nature to Live video: others on Facebook were experiencing the content at the same time, adding a level of empathy that might not otherwise be there.
  2. No matter what the format, our short form videos outperformed all our full-length Facebook episodes. Our short form series “One Word or Less” focused on explaining the concepts and different perspectives up for debate that week. Without exception, these videos performed higher in the most important metric we set out for our Facebook content: comments. On average, our explainers averaged 75 more comments than our full length episodes — and had three times as many views.
  3. Posting the full episodes allowed us to have localized conversations. By sharing the Live URL (crossposting during Live broadcasts is not yet available) with PBS and other public TV stations, markets across the country were able to have separate debates on their own Facebook pages. These conversations, we noticed, were very different depending on the market. We love the idea of hundreds of local mini-debates happening all around a central question.
  4. Fans have higher expectations for Facebook Live broadcasts. Once we started broadcasting the full episodes as Live streams, we noticed that people commented with the expectation that their comments or thoughts might make it into the conversation they were seeing. Although our broadcast workflow prevented this reality in Season 1 (we were live to tape, meaning we taped the show at 6pm and broadcast at 11pm), it opened our eyes to the possibility to be more responsive to a live audience online, not just from our studio audience.
  5. Viewers will stick around if you are clear with them what they’re watching. While we had a high retention rate of 32% of viewers watching past 10 seconds, the retention of our explainer videos was even higher — 38%. When working with long form video, the normal rules of social video don’t always apply. For instance, one of our concerns was the 15-second title sequence that came 30-seconds into our program — would Facebook viewers leave? Turns out, they were willing to stick around because we clearly defined before the break that they were getting the entire Point Taken debate experience.
  6. Timing isn’t everything. Much to our surprise, the difference in performance between posting at 11pm and the next morning was minimal. Our theory, based off the success of the Monday morning viral Facebook clips from Last Week Tonight with John Oliver, was that more of our fans are on Facebook between the hours of 7 and 10am than during the late-night 11pm time slot. We were also concerned that a low initial audience engagement in the evening would de-prioritize the post for the morning News Feed when our audiences spike. While this thinking may have been correct, the bigger factor at play trumped any concern over timing: the topic.
  7. Compelling topics were successful no matter the format. Our shows on technology, reparations, and prostitution took off on Facebook, enjoying an engagement rate 3x times higher than our other shows. Finding an issue that people care passionately about was the “it” factor for this show above any platform, format, or technology decision.

We learned a lot from this Facebook experiment and are thinking hard about the next season of Point Taken — what forms it might take and on what platforms it might resonate the most. As the ways society consumes and interacts with media shift, we want to remind people they can still have constructive and open-minded conversations on these new platforms.