Skip to contents
Hot Pod Insider

Apple’s In-Episode Analytics are out. What comes next.

Plus: A revolutionary new tool, The Daily goes direct, and more on the WNYC scandal.

Apple’s in-episode analytics feature is finally out to the public. You can find a pretty concise write-up by Recode’s Peter Kafka, but essentially, the new dashboard serves publishers a visual representation of completion and drop-off rates. From the new display, creators can also tell which sections were skipped — which means not only can publishers tell what segments under-performed compared to others in a given episode, networks can now also evaluate the presence of ad-skipping.

Obviously, this is a big change that’s been a long time coming, and I think I’ve spent enough time in this newsletter already illustrating the gravity of this new knowledge. (But in case you missed it, you should check out this past Tuesday’s last Hot Pod of 2017). A couple of notes to keep in mind:

  • As Kafka pointed out in this post, “The data Apple is providing only comes from people using its newest software: iOS 11 and iTunes 12.7. So it won’t be complete.” As of December 4, iOS11 adoption is up to 59% of devices, according to the App Store developer site.
  • Something I’m wondering: how will dynamic ad insertion impact the accuracy of the new in-episode analytics dashboard? After all, one of the notable side-effects of the tech is how it may lead to the different length deployments of the same episode, as different dynamically-inserted ads may not necessarily have the same length. I imagine this might further enforce length standardization among ads sold — so on so on midroll can only be 30 or 40 seconds, but the second midroll can be 50, etc. etc. — but I am curious how Apple’s tech itself accommodates for this.
  • This new feature matters nothing at all to podcast publishers whose majority listening happens off the Apple Podcast app, like Stratechery and Exponent’s Ben Thompson. Depending on where the listening of those shows might have migrated off to (say, Overcast), those publishers may already have had this kind of data for a while.
  • Speaking of Overcast, the app’s creator (and ATP host) Marco Arment posted screenshots of his show analytics on Twitter. Pretty nifty to sift through the responses there. [h/t Myke A]
  • Side note: I really really wanna know what people are finding! Rebecca Lavoie, of New Hampshire Public Radio and various indie podcasting adventures, has found that while the redesigned Apple Podcast App has possibly hurt her overall download numbers, she has found the findings of the new analytics to be thrilling. “My show is routinely over an hour. 90+ percent listen-through rate means our audience RULES,” she wrote me.
  • Something else I’m wondering: how will the new analytics impact the shape of podcasts itself over time? As NPR’s Rob Byers suggested: “I bet we see podcast lengths tighten up, if nothing else. Especially with the ‘one person behind a mic’ kind of podcasts.” Probably true, and I guess there’s also going be some trending towards being shorter. (For what it’s worth, I hope excessively long podcasts stay excessively long. I listen to way too many long NBA podcasts, and I love them so.)
  • But I think this is the more interesting question to ask: what will be the new value system that comes out of this new paradigm? Are we talking first-impression conversations that revolve around “60% of all listeners stay through to the very end” type of framing? To what extent will high completion podcasts — a space where shorter pods will win over longer pods — be valued more than the X percentage of listeners that make it to midroll six in a 1:45 episode?
  • A tangential question: we’re also likely going to some attempts to “game” the new analytics paradigm. How will they look like?

Anyway, that’s my scratchpad on the matter. And just for the record, I’m with Kafka when he writes: “There is considerable debate in podcastland about whether getting this data will be a good thing or a bad thing… I’m on Team Let’s Know More Not Less About The Things We Make.”

Don’t sleep on Descript. The new company, founded by former Groupon CEO (and creator of the audio tour platform Detour) Andrew Mason, is built around a technology that lets users edit audio in the same way that they can edit text. The platform essentially ingests audio files, runs it through a pretty good speech-to-text translation, and then throws the text up in the form of a document where users can efficiently cut words, tighten sentences, and switch around passages in such a way that will have corresponding effects on the audio file. It was initially developed as a creation tool for Detour, but in a fashion similar to Slack’s origin story, the platform was spun off into its own entity in pursuit of a whole different value proposition, which looks a whole like a potential pathway to compete with Adobe. The company has raised $5 million from Andreessen Horowitz.

In what will be the second linking to a Recode story in today’s newsletter, Mason guested Recode Decode this week to talk about the new company, and there are a couple of interesting things to take away:

  • On the arc: “We’re starting with people who need transcription, and then people who are working in the audio medium…  as we grow, we want to become the new standard of how talk-driven media gets made.”
  • On the underlying logic: “It just makes sense that you would be doing it from a script-driven viewpoint, as opposed to a timeline-driven viewpoint.”
  • Most notably: “We’ll be at a point within a relatively short period of time where you can produce a soup-to-nuts podcast in Descript.”

I’ve played around with the platform a little bit, and it’s radical just how many different small things could well be affected by adoption of this technology. Transcription has already been mentioned, but even something like the arduous (or, to some, meditative) task of tape logging could be infinitely tightened. There are also tons of questions to ask about how this might affect the accessibility of the medium itself; you could argue that it potentially makes it easier for more people to produce tighter audio, and that may have implications for more new people trying out the space with high levels of aesthetic ambition. (It feels a little reminiscent of the shift in video from analog to digital.)

That said, it all comes down to adoption, and adoption comes down to price point. After a free first 30 minutes, the platform charges 7 cents per minute of audio. That seems small, but that stuff adds up. And if you’re learning to create radio — which is a process that takes a lot of unintended time and tape — the small-adding-up might be a strong disincentive. And personally speaking, just in my own work as a person who listens to a lot of podcasts in order to write about them (with quotes and shit), I would love to use Descript to crank out transcripts for me to sort through, but my volumes aren’t going to work well for my budget. (My head seems more drawn towards the possibility of a Microsoft-style pay-monthly-for-tech-access model.) So, as it stands, even if the technology could be influential for increased accessibility to the medium, the pricing is where that dynamic is going to flesh itself out.

You can now listen to The Daily straight off the New York Times app. Which is really cool, and I wanna make a few points:

  • First of all, shout out to Bleacher Report, which did the podcast-listening-off-native-app first.
  • Aside from the obvious benefits in potential additional analytics (presumably off app interactions) and possibility for pushing listeners further down the subscription funnel, I think the main value is simply being able to serve an audience that’s already engaged with the New York Times brand but may have found it too much of a hassle to figure out this how-to-listen-to-a-podcast business, thus growing the overall listening pie. In my mind, making it easier for existing Daily listeners is of secondary value here.
  • This is why I think the move here is significantly different — and perhaps more valuable — than what NPR One conceivably does for the public radio system. This move to streamline audio consumption in app originally designed for other kinds of interactions automatically widens the scope for who The Daily can touch; NPR One, by contrast, doesn’t have that pathway of executing a similarly horizontal gain, and its value prop is primarily for people who are already bought into the notion of a more engaged audio relationship with the system. (Side note: I was unnecessarily snarky on Twitter when publicly wondering when the public radio system will make a similar jump. When thinking about it further, such a jump is not possible for NPR. They are simply fighting different fights. My bad.)
  • Worth noting: this is the opposite of the kind of fragmentation play by platforms like Stitcher Premium and Audible, whose strategic goal is to bring more people into its paywalled audio universe. The Times’ has a whole different paywall conversion funnel, and the switch here is that the (freely available, high quality) audio product is the bait.

Okay, moving on.

If you’re tracking the WNYC story… here are two stories, one from WNYC News and one from Village Voice, about the board meeting that took place yesterday.

Two more things on this:

  • As far as WNYC and the broader public radio system goes, the sexual harassment and misconduct stories cannot be divorced from the reality of pay inequities across the system. WNYC is an extreme expression of this deleterious dynamic; it’s a public radio station that has grown tremendously and gained tremendous power over the decades — breaking away as a city institution and becoming a global content force — but it has done so on the work system that is considerably dependent on a low-powered and unprotected of poorly paid interns and permalancers.
  • Further, this is a work system that has, for a long time, existed in an atmosphere of relatively little desirable competition. Mary Wilson, a producer on Slate’s The Gist, makes an extremely crucial point illustrating this on Twitter: “Without the podcasting boom, the great reckoning at @WNYC & NYPR does not happen. Podcasting blew open the market and gave women with stories to tell someplace else to go to work. Now you can speak up without ending your career.”

We all benefit from a competitive, decentralized economy in so many more ways than just merely economic.

Related story: Eric Weinberger, the president of the Bill Simmons Media Group, has been suspended over sexual harassment claims from his time at the NFL network. Bloomberg broke the story, and you can find it here.


  • After facing blistering backlash from its community, Patreon has reversed its decision to implement a controversial new fee structure. “Many of you lost patrons, and you lost income. No apology will make up for that, but nevertheless, I’m sorry,” CEO Jack Conte wrote in a blog post. I thought this Washington Post article was useful.
  • Dan Jeselsohn, formerly WNYC’s Director of Streaming and On-Demand Media Architecture, has moved over to Art19 where he will serve as VP of Digital Technology. (LinkedIn)
  • Gastropod’s Nicola Twilley and Cynthia Graber were interviewed by AdWeek about their indie operation, growing a show, and their Share-a-Thon campaign. (AdWeek) I first wrote about the Share-a-Thon in November.
  • ‘Tis the season of Nieman Lab Predictions!!!!! (Nieman Lab)