How to get another billion podcast listeners

In South Africa, Vodacom has around 45% of the cellular market.

Worked out using Vodacom’s R99 per GB “pay as you go” price as a percentage of the average wage (3.6x less than the US), the cost of data is stunningly high.

Mobile data costs a typical South African the equivalent of US $23.50 per GB of data.

Or, to put it another way, download the latest episode of This American Life and the 66MB for the episode will cost you the equivalent of $3.52.

This American Life isn’t a particularly data-thirsty podcast. It’s just over an hour long, and it’s a pretty standard 128kbps stereo MP3 file. It sounds nice.

Turn it into mono at 64kbps, and the quality doesn’t noticeably change. But that brings its cost down to $1.76.

MP3 is quite old technology for audio compression. Newer audio codecs like AAC-HE and Opus work better at lower bitrates. Turn This American Life into an AAC-HE file at 16kbps, the quality clearly reduces, but it’s still very listenable, and most importantly, it’s an eighth of the price, at $0.44 rather than $3.52.

Now, obviously, This American Life shouldn’t be encoding their audio at 16kbps for people in the US. Data is plentiful in the US, with free wifi all over the place and all kinds of unlimited data plans. But that’s not the case in other countries.

But what would happen if podcasters, or apps, were to offer additional lower bitrate streams?

Many years ago, Stitcher, a podcast app, used to transcode most podcasts down to something like 48kbps. It was meant to allow you to use it in the car: where 128kbps or 192kbps reception wasn’t always guaranteed. Podcasters complained and mobile networks improved, so eventually they stopped.

Google Play Music, too, used to cache podcasts and transcode them where necessary to allow listeners to enjoy the podcast where a high bitrate was difficult to receive. It’s now closed; and Google Podcasts doesn’t cache audio.

Both of these services did transcoding and caching because there wasn’t a way to offer lower bitrate versions of the audio. (Unless you started a whole new RSS feed, which makes things confusing for everyone).

They worked in a very Western way: choosing a lower bitrate version if the connection wasn’t good enough (irrespective of whether you could actually afford the data).

I think we can, and should, do better.

One of the new proposed tags at Podcast Index is the alternateEnclosure tag, which is partially designed to offer alternative ways of downloading audio (through things like IPFS or torrents), but is also designed to offer, literally, alternate enclosures at lower bitrates.

This would allow me to publish a very low bitrate version of my audio, for those who want it. It allows for different audio codecs, like this version of Podnews in the Opus format (at 32kbps). And it’s backward-compatible: podcast apps that don’t understand it would just take the standard enclosure as normal.

A podcast app could know whether it was on cellular, and choose the lowest supported bitrate direct from the podcaster.

Or, perhaps, a podcaster might be able to suggest “yes, it’s fine if you cache and transcode my audio” in the RSS feed as an alternative. That’s not written into any specification, but could be: simply an additional value within the main enclosure tag, perhaps reading podcast:cacheable="yes" .

Both of these things could enable a new generation of podcast apps who market themselves as offering podcasts without the data costs. And, by extension, many new listeners.

There are billions of people out there who cannot afford the data for their phone. Podcasting should be inclusive: not just for those behind the microphone, but also for those who want to listen.

Would you make a low-bitrate version of your podcast available using a new alternateEnclosure field?