Pop Up Labs: Sharing audio on social networks
What’s holding podcasting back from broader adoption? One problem is that podcast listening is often siloed within podcast apps, without direct connections to broader social networks. Finding easier ways to share audio where people are already hanging out — on social media — can help attract new listeners.
To help tackle this problem, the Pop Up team has been experimenting with audio players for Twitter. Our Audiosear.ch podcast engine (more detail in case you missed it) generates audio players that open right from the feed in the Twitter desktop website and app. That means any episode link you tweet on Audiosear.ch will include audio.
You can tweet specific moments, like this audacious clip from The New Yorker Out Loud Podcast (shown on right).
With help from producers at Buzzfeed Audio, Panoply, and Gimlet Media, we’re tracking data about how Twitter users interact with audio. We’re counting things like how many Twitter users hit “play,” how long they listen, and what they’re most likely to listen to. Here are some of our preliminary findings:
Does the length of the audio clip matter?
From the audio we’ve tracked so far, people listen to more of the audio for shorter clips compared to longer clips. For audio that’s less than two minutes, users listened to an average of 89% of the audio, compared to 39% for audio over two minutes.
More click “Subscribe” than “Follow,” but no one’s doing much of either
Once people listen to audio on Twitter, we want them to be able to jump straight to the creator of the audio: the producer. We customized our player and A/B tested two different buttons: “Subscribe to [Show]” and “Follow [Show].” Our hypothesis was that users might be more inclined to engage with show links if they were framed in social media language, rather than “Subscribe,” which might feel like more of a commitment.
So far, our results have shown a statistically significant preference for “Subscribe” over “Follow.”
It’s hard to separate form from content
As we saw from the high engagement for the Hillary Clinton audio clips, it’s hard to tease out what patterns are the effects of technology versus to the compelling nature of the content. Our goal is to collect enough data that we can normalize for form and content to identify what makes the best player AND the most shareable audio.
Stay tuned as we continue to develop the player and track more Twitter audio data.
In the meantime, have any audio you’ve been wanting to share on Twitter? See the Twitter player for yourself by tweeting an episode or clip from Audiosear.ch.