I don’t have the energy for a lengthy article so forgive me for jotting down a few quick thoughts about Luminary and paid subscriptions:
For podcasters, more ways to make money I think is a good thing. If you think about every other media type there are multiple ways to earn money. If I want to watch a movie I can buy it, rent it, go to the theater, wait for it come to TV with ads. Why must podcasts only be ad-supported? Luminary is focused on the top-tier glossy productions but starting to get people used to the idea that they can pay for podcasts is beneficial to everyone.
For listeners, fewer ads is great. I would be very happy to pay money to never have to hear a podcaster talking about their anti-microbial underwear again. Of course, Luminary is not offering all podcasts ad-free, but you get the idea. Side note: I wonder if we’ll see more incubators and creator programs from Luminary to help podcasters develop their ideas and persuade them to move to their platform.
Going in with a bunch of celebrity-led shows was really the only option. Movie execs talk about stories with “pre-built audiences” (think adaptations from comics or podcasts) and this is no different. By paying for a celebrity host, they are really buying that celebrity’s audience. For this to work long term, the content needs to also be good but the pull of star power is the how they hope to build their audience to being with.
Offering all podcasts is another way to draw people in. I was surprised to learn that Luminary was also a regular podcast player. But it makes sense; if they can offer a superior listening experience for all podcasts (maybe that is what the 40+ engineers have been working on?), then they could get listeners to switch to listening on Luminary and you better believe they will be promoting the heck out of their paid exclusives to anyone using the app without being a subscriber.
So I guess I’m feeling reasonably positive about all this?
On February 27th Apple sent out an email to apparently everyone who has submitted a show via Podcast Connect because Apple wanted to clarify how some of the iTunes tags should be used.
There was then some kerfuffle around using episode numbers in titles that Apple then clarified wouldn’t result in your podcast being removed. But I want to talk about the third bullet point, “Incorporating irrelevant content or spam”. Apple is asking creators to stop “keyword stuffing”, the SEO practice from the late 90’s where a creator would add as many keywords that they thought people would be searching for so that Google would surface that content to those readers. That approach was short-lived because instead of asking people what the keywords should be, Google could just index the actual document itself and use that to determine how relevant it was.
While I agree that keyword spam is a problem, just telling people to stop doesn’t address the problem of why people are doing it in the first place.
When there are very few things a creator can do to get their podcast discovered, other than setting a category and asking people for a 5-star review, their only option is to try and game the search system by adding as many keywords as they can to episode titles and descriptions. If I search for a topic in Apple Podcasts, guess what? Shows and episodes at the top of the results all have that phrase in their title. On one hand Apple says, don’t do keyword stuffing and yet they encourage that exact behavior on their consumer app.
Instead of just telling creators to stop, you need to give them an alternative way to show that if they create good content it can be found by listeners. This reminds me of Marco Arment’s story about trash can positioning in the men’s room at Tumblr where someone wanted to stop people dropping paper towels next to the door. The first attempt was to put up a sign saying “stop dropping paper towels”. Apple’s email was someone putting up that sign. What they should be doing is addressing the actual problem and move the trash can next to the door.
How can they do that? Well, they bought Pop Up Archive and their tool Audiosear.ch in 2016 whose whole mission was to index episode content and extract topics and meaning. Apple owns a tool that solves this exact problem! Of course, shipping a new feature is far more difficult than sending a clumsily-worded email, but I do hope that one day we see the fruits of Pop Up Archive acquisition.
Oh great question that I totally didn't address! It would be fascinating if they didn't... But yeah I think they will, the app paradigm is too strong and I don't think voice is going to be the dominant UI for a little while yet...
As reported by 9to5Google, some translation strings have been found in the codebase for Android 8.7 that strongly suggest that they have a dedicated podcast app in the works.
Previously Google’s podcast strategy was to include podcasts along side music in the Google Play Music app so it will be interesting to see how well a stand alone app is received. I imagine Pandora and Spotify will be following the results closely.
There wasn’t much podcast related news at WWDC this year. The headline figure is that Apple now has 555,000 active shows (of course it is still unclear how “active” is defined) which is up from 525,000 in April. I think they might have Anchor to thank for that boost in numbers. TechCrunch has a decent write up of the rest of the news.
There is one thing that doesn’t seem to have been widely reported that I think is the most interesting thing announced. With watchOS 5, third party developers can now play background audio on apps running on the Apple Watch. Which is a huge deal for podcast apps that aren’t made by Apple. Previously the APIs offered to third party developers were not enough to create even a half decent playback experience on the Watch (which is why you don’t see a Pandora or Spotify Watch app). For those who want to understand a little more, Marco Arment has detailed the reasons why building a podcast player on the phone was impossible with the previous versions. Now watchOS 5 addresses most of those concerns. It will be interesting to see if people want to use their watch for playing podcasts.
RadioPublic has continued the trend of episodeunbundling with their latest feature: Stations. The idea is that you can provide a better discovery experience by grouping together episodes across a similar topic but from multiple podcasts. For example, one Station is called ‘Sports Discussion - reflect on recent games while at the gym’ which collects episodes from podcasts such as ‘Pardon My Take’, ‘The Bill Simmons Podcast’ and ‘First Take’.
RadioPublic has offered a similar feature in the past called Podcast Playlists. With Playlists such as “New to podcasts?” that also grouped episodes into playlists based on similar topics (this feature now seems to have been removed from the iOS app at the time of writing). What makes Stations more exciting than Podcast Playlists is that while Podcast Playlists were a static collection of episodes, Stations will update automatically based on what you’ve listened to:
…we want to assure that when you press play on a station, you get content that’s right for you. By design, we’ll respect the show’s natural play order, if it’s a daily or an episodic show, we’ll play you the most recent (unless you’ve heard it, of course) or if it’s a serialized show, you’ll start at episode one (or pick up where you left off). The station will continue to cycle through all the shows on the station to always serve you fresh audio.
Podcasting hasn’t had a great ‘lean-back’ experience yet (that is, a listener can ‘just play something I like’ rather than choose, download, queue and play a specific episode) and my first impressions of Stations is that it is a great lean-back experience. And they have cleverly given a cue to listeners about when they might want to try each Station by giving them subtitles such as “Commute while catching up on current events” or “Eat lunch with a legend”.
Who is Stations for? I can think of three main use cases:
A new listener who doesn’t know where to start but they know they like sports news
A enthusiast listener who has listened to all their shows that week and is on the hunt for something new
A social setting with more than one person. This one I’m particularly excited about; you often hear how podcasting ‘isn’t a social experience’ and I think that is true for two reasons: its difficult to just have audio as the only focus of a group situations and there is a lot of pressure on the person who plays the podcast to choose the right thing. Stations offers a way to play podcasts without the pressure of having to be the ‘DJ’ for the group.
Are there any problems though? I do wonder if podcast publishers have any concerns about Stations? Maybe it could be argued that episode unbundling is similar to what has happened to online publishing. Now it is no longer about the publication as a whole but about individual articles and who can make the most clickbait headlines for going viral on Facebook. But I don’t subscribe to that argument. With editorial curation this feels like a system for surfacing great content rather than just being about gaming an algorithm for short-term popularity.
But might there be other problems for publishers that I’m not considering?
Charts are always going to be gamed. Anything that is essentially free publicity is going to attract those who want to exploit that for their own benefit. Publicists and journalists are often eager for any corroborating evidence to pad out their pitches and articles. Of course, for podcasts there is no central source for this kind of data. Apple has long been the dominant player and the Apple’s podcast charts have pretty much been the only stats in town that anyone can point to. (As an aside, I was pondering why they chose ‘hotness’ over just pure downloads for their chart? My guess is that they want to avoid the top of the chart being stale and to surface shows that have just launched but might not have huge download numbers yet. Keep things interesting y’know?)
It feels like there is an opportunity for someone to make a new chart based on downloads and then something like a ‘viral 50’ too. Spotify or Pandora or Google could provide an independent (-ish) chart easily. In fact, they could do it based on actual listening rather than downloads! But of course they don’t have very large podcast catalogs right now. And while being a part of the conversation around new shows and how they perform could be important, there might not be a clear correlation with making a podcast chart and increasing listeners that would warrant the expense of building out a chart system.
Something else I’ve been mulling over, what about episode-level charts? Forget podcasts, what are the most downloaded episodes this week? Would that be interesting for creators and listeners? What do you think the most downloaded episode of all time would be?
David Murphy, writing for Lifehacker, reports on his experience using the new Android experience for downloading and managing podcasts. Other descriptors used in the above article include ‘crude’ and ‘curious’. But I think that is doing a disservice to Google and shows how we expect everything to be an app nowadays and we get confused at things that blur that line.
One of the many clever things about the iPhone was the app-ification (yuck) of every function on the phone. To make a phone call, you used an app .This meant that making a phone call was just one function of the mini computer you were carrying. Nowadays we assume everything must be an app, which might explain why Mr Murphy felt Google’s podcast experience was ‘curious’. Google has opted to not make a podcasting app but instead extract its functions into the web browser and the operating system. The weakness of the app paradigm is that the user needs to know which app does what and where to find it. To listen to a podcast, the user needs to know what a podcast is, know where to find the app and know to open it and how to search for a podcast once there are there. If you move away from that, you might be able to provide more continuity of service if you can merge searching, playback, sharing and surface podcasts as an option when people are searching for specific information.
The other thing to consider is that as voice controls and smart speakers become more prevalent, we move away from an app-centric (yuck again) experience. If you ask your Google Home to play NPR, do you care if it comes from NPR One or Spotify or the Google podcast app? With voice I think we move away from the idea that there is a ‘podcast app’ and towards the idea that it is just ‘Google playing my podcast’ and seamlessly integrates and syncs with my computer, phone and smart speaker. And this is key if Google are going to ‘double the audience’ for podcasting (as they have stated). They aren’t going to hit that goal by explaining how to use an app. They are going to get there by offering the most convenient way to listen to people who aren’t already listening.