Day 1 of #productidea — Improving the experience of listening to podcasts
This is part of my #productidea 30 day challenge on which each day I write about a new product idea.
Let me start with a confession: I don’t usually listen to podcasts. I’ve tried several times over the years and I always end up feeling that I’m wasting my time. I don’t like spending 1 hour of my time for just 10 interesting minutes. Each time that I try again I keep finding myself on that position. Maybe I’m not made for podcasts, or maybe I haven’t found the right type of podcast for me.
I’ve been thinking if there could be a better experience for me.
Discover the best episodes
Let’s suppose I am new to podcasts and I check out the Tim Ferriss podcast because it’s very popular. I see that he has done over 150 interviews. In order to check if I like it, should I start with the latest episode or the first one? Or should I try to find one with a description that catches my attention?
None of those are the best on-boarding for a new listener. The probability that one of those are one of the best episodes is very limited. The experience would be definitively better if I start with one of the most liked episodes. I’m pretty sure Tim would gain more engaged listeners that way. Of course, that only applies to podcasts with independent episodes (Serial doesn’t).
Planet Money is a great podcast, but some episodes are more captivating than others. If I’m not a hard-core fan who listens to all their episodes, the alternative is to listen to none. There is no middle ground. Either you’re subscribed or you’ll forget to check their new episodes frequently.
Let’s say I have 4 hours a week when I can listen to podcasts. Am I better sticking with my regular podcast episodes or trying a new one? How much time should I concede to an episode if I’m not liking the first minutes before dismissing it? Time is limited so I should be choosing wisely.
Nowadays podcasts apps focus their recommendations on popular shows but I believe they should start focusing on recommending episodes to address the previous problems. And I’m not the only one thinking that (thanks Anna for the link).
- Tell me which are the most popular episodes of a podcast, so I can check if I may like the show.
- Tell me which are the most popular episodes I haven’t seen of the podcasts I follow, so I can optimize my 4 hour span.
- Tell me which are the most popular episodes of podcasts I don’t know, so I give the best shot at finding new podcasts.
- If I want to improve my productivity, tell me which are the most popular episodes on that topic; giving me three podcasts with 500 episodes each doesn’t help.
The problem here is capturing the data. There are two interesting pieces of data: one is passive (has the user listened to 10 or 90% of the episode?) and the other is active (the user recommends the episode). Who has this data? The podcast apps.
Apple could build this inside its podcast app, but we all know they suck at recommendation engines; so I wouldn’t put my hopes on them. Another bet could be Marco Arment with his Overcast app. It may have enough active users to get relevant results. In fact, I bet he is already planning something. On the Android ecosystem, they will have their equivalents.
Read the summary of episodes
Sometimes after listening to an one hour podcast, I feel like it would have been more efficient to have read a 5 minutes article. Of course, podcasts are not about efficient, it’s another medium to transmit stories and information in a more relaxed away.
Even so, I think there is an opportunity to take the key takeaways of the conversations and capture them in a summary so podcast listeners could decide if it’s an episode worth listening for them or not. This may sound counterproductive for the podcaster as he wants more listeners, not less. But I believe it’s also a sign of respect for his listeners.
If I like what I read in the summary, I’ll listen to the episode. If not, I will wait for the next one. But I won’t have that sense of frustration of having lost my time.
You can find an example of this approach on The Podcast Wire, a weekly newsletter with summaries of the podcasts they recommend more. I’ve already subscribed, I’ve glanced over several summaries and I will listen to two episodes whose summaries I found really interesting.
Listen to pieces of an episode instead of the full hour
There are other times, specially on the interviews, on which I am only interested on a specific topic; not on the whole conversation. Let’s say I want to listen to everything regarding the Apple vs FBI case or the news about VR in the Mobile World Congress. I want to search about this topic and just listen to the minutes that talk about it.
Some podcasters add marks to their episodes so you can advance directly to the topics that interest you; but it’s not common. Without their support, this task is more complex as doing it manually is really time consuming and I’m not sure technology allows us to do it cheaply automatically yet. I’ve been playing with IBM Watson voice recognition and it’s quite good getting the text but then we would need to detect topics of conversation, not just words. But it’s true that once we have the text, we can apply the same algorithms that are currently being used to group stories by topic.
I’d love to know your thoughts.
