abstraction
killing it with close inspection

Two Timelines for the Future of Podcasting

On January 6th, 2019, Spotify announced its acquisition of two podcasting startups: Gimlet Media, and Anchor. These acquisitions indicate a broader shift in the business landscape of podcasting.

Podcasting is a historically democratic medium. Shows are hosted decentrally, and distributed to clients via RSS; anyone may host a show, and anyone with a client capable of parsing standardized podcast RSS feeds may listen. This decentralized model has its disadvantages. Without a centralized, browsable repository, podcast discovery would be too ad hoc to initiate or sustain substantial growth of the medium. Apple's podcast repository has become the de facto industry standard, providing unfettered access to a policed list of podcast feeds numbering over 500,000 in 2018. Still, Apple does not host any of the shows in its library. Though the industry remains decentralized, without Apple's - thus far - benign centralization of podcast feeds it would surely stagnate.

This model is a net good for the still-nascent industry. Decentralization of hosting allows creators to retain control over their shows, and the centralized de facto source-of-truth for podcast discovery has provided the stable infrastructure upon which podcasting has been commoditized. The default revenue model for podcasts remains no revenue model; a significant majority of podcasts make no money for their creators. Still, a growing number of podcasts have opened their doors to advertisers. These creators have embraced the data-poor nature of the medium, selling ads against a estimated audience size and the demographic information inferable from the topic of their show.

Neat.fm, for example, manages ad sales for two shows, The Accidental Tech Podcast, and John Gruber's The Talk Show. Advertising inquiries are plausibly made on the basis of the information provided on the website, limited to approximate stats, a brief description of the show, and the price of an ad-spot. There is presently little analytic transparency to podcast advertisements. Advertisers can't tell if they're ads have been skipped, and in which demographics they have lead to sales. An even more fundamental question, the attribution of sales to particular ad spots or campaigns, is also difficult to answer. Clever advertisers include a campaign-specific discount code to make this attribution. Crucially, this opacity exists _because_ the medium is decentralized.

Imagine, if you will, a world in which podcast distribution is centralized. Distribution of episodes no longer takes place over RSS, and the audio files are hosted by large companies. Proprietary content distribution APIs mean that feeds can no longer be parsed by just _any_ podcatcher. The proprietary podcatchers that remain (if they deserve the name at all) will phone home with data about the user's interests, listening habits, and propensity for skipping advertisements. Once enough data is amassed, advertisers can begin to dynamically insert relevant advertisements into episodes of your favourite show - remember, they host the files now and can exploit file-streaming protocols to splice advertisements into the delivered audio. Soon, these streaming companies will look to exclusive shows a means of differentiation, enticing listeners into their walled gardens with promises of celebrity hosts and unmissable audio-dramas. The once-decentralized podcasting industry will be fragmented, and shows which have remained independent will eventually succumb to a combination of listenership attrition and the flight of advertisers to data-rich platforms. The promise of free, host-anywhere-listen-anywhere audio will have given way to siloed, subscription-based audio distribution.

And on January 6th, 2019, Gimlet Media and Anchor made that future all but certain. In Gimlet Spotify gains a content-production powerhouse capable of generating exciting, exclusive shows for the platform. Gimlet has already produced its first Spotify exclusive show, _Mogul_. (Spotify has recently acquired Parcast, and partnered with the Obama’s Higher Ground Productions, bolstering their catalogue of exclusives) Anchor's value proposition is no more mysterious. Their platform provides a barrier-free entry to podcasting, hosting audio for free and matching podcasters with advertisers. For now, Gimlet still distributes the majority of their shows via unrestricted RSS feeds, and Anchor still allows publishing to the same. One day, Spotify may choose to centralize much of podcasting with the flip of a switch.

The argument might be made that Spotify's investments in podcasting will lead to more high-quality shows. And this might well be true. But podcasting was thriving before Spotify's intervention, and would have continued to thrive without it.Alternative revenue models for podcasts exist. Some shows have banded together into networks, selling ads against the size of their collective audience, or attracting celebrity hosts of their own. Networks, like Relay.fm and Maximum Fun, often accept direct, recurring  contributions from their audiences. Individually popular shows like Do By Friday might turn to Patreon for additional funds. Community and philanthropy are also enabled by podcast networks, and Relay.fm recently raised USD $314,269 for the St. Jude Children's Hospital in Memphis.Luminary, too, has tried to enter the arena of premium podcasting, enticing high-profile celebrity hosts with mountains of venture capital; for CAD $8.49/month, you, too, can listen to a podcast from Trevor Noah.

But celebrities are _already_ podcasting. To name a few: Rami Malek starred in audio drama Blackout, sponsored by Sonos, _David Tennant Does A Podcast With_ and _Conan O'Brien Needs a Friend_ attract weekly celebrity guests. Advertising dollars _are_ being spent on celebrities in sufficient quantities to attract them to the medium. And celebrities often don't _need_ the ad dollars, either, choosing, like Arlene Dickinson and Jan Arden, to eschew sponsorship in favour of greater latitude to speak on - often deeply personal - topics of choice. The stratification of the medium will, no doubt, encourage advertisers, but it threatens the viability of conventional podcasting.

Alex Blumberg and Matthew Lieber, co-founders of Gimlet Media, identified a disconnect between the number of people listening to shows and the amount of money coming in from advertisers. They're solution was to abet the destruction of the medium they helped popularize. Sure, on-demand audio shows will exist, but they will scarcely deserve to be called podcasts. Lieber and Blumberg were not the first to sell out their industry, but their decision feels personal, and their rationalization entirely misses the point.