Podcast Episode Management For Tomorrow

Read The article

A new client of mine has 3,700 episodes of their podcast. They do an episode every day, five days a week, and have for about five years. Do the math. It's about 3,700 episodes. This client has contracted with my firm to help migrate to a more modern hosting environment. While the technical implications are interesting enough, it’s giving us a chance to really think through the strategy of presenting those files to potential listeners.

We know there’s merit in having all 3,700 episodes available on the website, so we’ll migrate them all and update all the media file links. But we’re thinking about the in-app listening experience for that extensive of a catalog. That’s more worrisome for a catalog that big.

Today, podcast subscribers only have two options when it comes to consuming available episodes: start from the most recent episode and work backward, or start from the very first episode and work forward.

Neither of those is very attractive for 3,700 episodes.

Even though Apple gifted us with the ability to break up batches of episodes into seasons, listeners’ options are still limited for getting all the episodes within a given season. Yes, that’s fewer episodes to rifle through. But the listener will have to repeat the process multiple times to evaluate the full show. Is that helping? Or just delaying?

In Which Evo Once Again Blames Podcasting’s Woes On Blogging

Podcasting inherited this paradigm of date-based sorting from blogs. The reverse chronological structure of RSS feeds is well-suited for syndicating new items to subscribers. You publish a new episode, your hosting company (or website) updates your show’s RSS feed, and all your subscribers get it. That process relies on date-sorted content to work properly.

We also inherited a distributed architecture… and then fractured it even more. You host your podcast files on your chosen podcast media host. I host my podcast episode files on a totally different podcast media host. And we probably also have a podcast website in the mix that’s hosted on a completely different set of servers.

While that’s great for resilience, it inhibits the development of a solid recommendation engine ala YouTube’s recommendation engine. Neither download nor consumption data of episodes is shared between hosting companies or listening apps, so that’s a big hump to get over.

Podcasters Taking Matter Into Their Own Smart Hands

Some podcast owners are taking the initiative of re-grouping and sorting episodes on their podcast’s website. I think that’s a smart play and know if a few data-minded people looking at web analytics, media hosting company downloads, and consumption data (like from Apple Podcasts and Spotify) to make a rudimentary recommendation engine of their own content.

But none of those efforts matter at all once someone becomes a subscriber. Once they’ve subscribed, they’re off your website and are back to two options: listen backward or listen forward. There’s no connection between the episode they just listened to and another episode in your feed. Except for the one that came immediately before it.

Sure, the listener can “click out” to get to the webpage for the episode where that savvy podcaster has created a list of next episodes to listen to. But now the listener is no longer in their app, and wasn’t that where you wanted them in the first place? How do they easily navigate back to their chosen app once they’ve received a recommendation from your website?

It’s messy, even if you don’t have a thousand episodes. This is the 305th official episode of this podcast. Add in the bonus episodes and that’s closer to 320. Yes, some people do download the entire feed and start from episode 1. Others just work their way backward from wherever they start. But neither is really conducive to enjoyable listening. What they all want, even if they don’t know it, is to be guided through the episodes they want/need to hear about.

There has to be a better way to present this. Temporal alignment can’t be the only way. But it’s the only way I’m aware of today.

It’s easy to leave the burden on the shoulders of the creators. After all, we made each episode. If anyone knows which episode someone should listen to next, it’s us. But again, we don't have a good architecture to manage or present this. While I might be able to go through all 320 episodes and put in better tagging, categorization, and connections, my efforts are quickly out of date.

It’s A Podcast, Not A Magazine Or Newspaper

For some podcasts, sorting by date is perfectly fine. Some episodes are captured moments in time, so the date alone might be enough to see what one podcaster had to say on July 29th, 2007. Maybe that was an important day in history and understanding what was said on that day is worthy of examining, much like you would a magazine or a newspaper. In that case, the date-sorting we have today is perfect.

But I bet if I went over to your house and I looked at the books on your bookshelf, I would not find your books organized by publication date.

Figuring out a better way that works for all podcasters hurts my brain.

Will Smart Earbuds Solve The Problem?

I imagine a future where smart earbuds change the game. Thanks to the growing adoption of smart speakers, we’re becoming more accustomed to using voice assistants during our day. These voice assistants listen to us. Perhaps they can listen to the episodes we’re enjoying, stepping in when one has ended and making a recommendation of what we should listen to next?

While I can barely remember what my episodes were about last week, AI-powered tech should be able to store contextual information on a much larger scale. That’s the direction we think Google is taking with its integration with Google Search. And Google’s voice search continues to improve. So it doesn’t take a crystal ball or corporate espionage to see where that’s headed.

Curmudgeons may disagree and privacy wonks will sound the klaxons, but a future where deciding on what episode to listen to next is truly a hands-off experience guided by a smart device is rather interesting. And a bit obvious.

Speaking of interesting futures, I’m visualizing you, right now, going to BuyMeACoffee.com/EvoTerra and, you know, buying me a coffee.

I’m also envisioning you making a phone call or sending an email to one podcaster you know and telling them about Podcast Pontifications. Word-of-mouth efforts are really what spreads an ultra-niche show like this, so thank you in advance.

I've got all sorts of weird, crazy things on my brain, so that's what I'll be covering all week long. I'll see you tomorrow for another Podcast Pontifications.

Cheers!

Download The Audio FilE

Display/Hide Transcript

PP305 - Podcast Episode Management For Tomorrow

Evo Terra: [00:00:00] Say that a new listener discovers an episode of your podcast and loves it. Now they want more, but they don't really want to or need to run backwards through your entire list of 300 episodes, but sadly, that's all they can do today.

[00:00:23] Hello, and welcome to another podcast. Pontifications with me, Evo. Tara. I was chatting with a client over the, well, towards the end of last week, they have 3,700 episodes of their podcast, 3,700 and that's because they do an episode five days a week and half for 15 years. Ish do the math. It's about 3,700 episodes.

[00:00:48] Wow. That's a whole lot. And we're moving to another hosting platform. When we were thinking, do we need to bring all this other stuff. You know, his question to me was, what's the likelihood that someone's going to subscribe to our podcast? And even if they did download all 3,700 episodes, what's the chance they'd go through and listen?

[00:01:07] That's a lot. Even if the episodes are 10 minutes long, that's 3000 at 37,000 minutes.

[00:01:15] Music: [00:01:15] That's a lot.

[00:01:16] Evo Terra: [00:01:16] What's the chance of that happening? And I said, well, probably none, but we need to think about things. Because what we need to think about is there's more to just what's in the feed and what people actually consume, right?

[00:01:29] Feeds are limited by all sorts of things, and again, I think that going backwards through 3,700 episodes is not really a pleasant time, nor is the prospect of downloading 3,700 episodes and going back to the beginning and starting with episode number one. But that's really all we can do today. Okay.

[00:01:48] Thanks Apple. You gave us the season tag so we could download those in seasons, but still our options are limited to listen backwards or listen forwards, whatever that means. Listen from the beginning of the feed, way back in time, or listen to the current feed and go backwards and that's not great. Now we inherited this paradigm of date sorting by default from blogging, our shared common ancestor, or maybe we say Bootsy, people would say we evolved out of blogging, whatever, but the reverse chronological sorts that we have in podcast feeds is completely driven by blogs and the way RSS feeds work.

[00:02:32] Most recent updates, so therefore date sorted. And our distributed architecture, which means you host your podcast files on your podcast media host. I host my podcast episode files, maybe on a totally different podcast media host, and they're the two shall meet. That also inhibits something other than that, because.

[00:02:57] Well, we can't really put together a solid recommendation engine when content is hosted across multiple hosts who don't share data on consumption back and forth with one another. You tubers have it nice because the YouTube algorithm can see what you're listening to and make recommendations of other things because they host all of the content on their own.

[00:03:19] We don't have that in podcasting. So while you might be able to, as a podcast host or a podcast hosting company, or maybe just a website owner, you could probably do some advanced analytics of your own content to determine what kinds of things are popular. Maybe do some resorting, but that may not work at all for the vast majority of people who listen to your show because they use a podcast app and at the end of an episode that they're really enthralled with, they have two options.

[00:03:50] Keep going to the next episode or see a list of all episodes. That's it. Okay. Maybe clicks on a website, but you know, in app there's not a lot of options that people have. How do they navigate from spot to spot? I've struggled with this. I just hit 300 this is the 305th official episode of this show.

[00:04:13] Added of the bonus episodes, probably closer to 320. What I want people to do. Go back to the beginning. No, I don't want them to do that. Although they do listen backwards. Just catch it. I don't know. It's just not conducive. If you like this episode, you might like, this episode doesn't really exist and I think it doesn't exist.

[00:04:35] Well, I know it doesn't cause this because of the things I just mentioned, right? We don't have a good architecture set up for this. Now, you could argue, as I have tried to argue myself into the idea of. Maybe I go back to those 305 320 episodes and do some better kegging tagging and categorization, but I don't necessarily think that's the case because that's set up for manipulation.

[00:04:59] All sorts of other problems. When you come, when you do manual tagging and stuff. I mean, sometimes podcast episodes have lots of different things that they put in them. Mine don't, mine's every episode is about one thing, but a lot of other podcasts aren't that way. Which is why for some podcasts, the reverse date order is perfectly fine.

[00:05:19] I want to know what this show was talking about on July 29th, 2007 okay. That might make sense. It might've been something in history worthy of talking about, but a lot of podcasts don't put out date based information. Database sorting is fine if you're a magazine or a newspaper, but. I bet if I went over to your house and I looked at the books on your bookshelf, they were not organized by publication date, I can almost promise you that's the case.

[00:05:54] Same thing in podcasting. There's gotta be a better way to do that. We need to be thinking about this as an industry, as players inside. How would we like. Our episodes to be presented to someone in an app. There isn't an app that does this today, at least not that I know of. And developing that is just hurts my brain thinking about it.

[00:06:22] But I do imagine a future, and maybe it's because of the smart earbuds that we have now and the voice control getting much better. People are being used to voice assistance. They're becoming more accustomed to voice assistance and utilizing them in their day. And voice assistants are helping out in lots of interesting ways.

[00:06:44] Maybe voice assistant, which is designed to listen to us, has enough technology in it to listen to the episodes and maybe the brains of the computers can make better recommendations. Maybe if you've listened to an episode and your smart earbuds also listened to that episode, maybe at the end, those smart earbuds can say to you, well, if you enjoyed this, here are three other episodes you might enjoy.

[00:07:16] That would be kind of interesting because that AI that's powering all back and hold a lot more information than we can. And I guess it's just a matter of, you know, listening to everything. To get that index built up, which is kind of where we think Google might be headed and maybe not with the AI can, although why not Google voice?

[00:07:38] He said quietly so it doesn't activate. Maybe that's what the future looks like. Maybe all this work we do on search engine optimization, which we've talked a lot about and other sorts of things. Maybe it's very different when the, this. New paradigm is put in place, and I hate to just overuse the word paradigm, and there's some intelligence helping us Lisette together.

[00:08:00] Wouldn't that be an interesting world where when you're done listening to an episode, it's a truly hands-off experience where the robot comes back on and helps you decide what your next piece should be? That's kind of interesting, which will make us change a lot of things. Yeah. I think that's an interesting future for us to imagine that speaking of interesting futures that we can imagine, I can imagine right now, you going to buy me a coffee.com/evo Tara and leaving, you know, buying me a coffee.

[00:08:37] That's what that's all designed to do. Even better than that. However, I can envision you. Picking up the phone, sending an email to one person who you know who's a podcaster and telling them that they should be listening to podcast pontifications.com your word of mouth efforts is really what makes this show, especially when that's weird and niche-y like this.

[00:09:00] It's the only way we can really grow. Do that for me. I would appreciate it. I've got all sorts of weird, crazy things on my brain, and that's what I'll be covering all week long. So I'll see you tomorrow for another podcast. Pontifications cheers.

‍

Comments