I have often puzzled over the relative ‘effort to create’ vs ‘time to consume’ ratios of podcasts vs an essay.

A 1,500-2000 word essay takes me nothing less than 6-8 hrs to create + publish, sometimes more if it needs background research etc. It take the reader 10-15 mins or so to consume it. Let us say, on average it takes 8 hrs to create and 15mins to consume. So 1:32 is the create-consume ratio here. I am probably not very efficient on the writing front, and yes, the writing time will change depending on the topic and type of essay. But I doubt if it will ever go below 1:10.

Contrast that with podcasts.  It takes roughly 4 times the podcast length to create an episode. This includes preparation, time to set it up, the coordination, the edit to clean up the podcast and ready it for publishing, and then the posting to multiple sites + social media work. So 1:4. This is the average ratio, and I sure there are folks who will be operating at 1:2 or 1:3 ratios even. Perhaps there are people who do a lot of background research and have to read books by their guests etc., so there the ratio could approximate that of writing, but the vast majority of podcasts are likely to be in the 1:4 create-consume ratio category, or thereabouts.

Different attention currencies

Now podcasts do take far more time to process than an essay. The average podcast length is 43 mins and the average time to read for the most recommended Medium essays was 7.3 mins. I know I am not comparing like to like but this is the best I could get. Roughly 6x more time to consume a podcast vs an article. That said, the 7 mins that you spend on reading an essay are very different from the 43 mins you spend listening to a podcast. The former demands a more intense kind of attention. The podcast on the other hand is something you listen to while doing other chores or activities – laundry dishwashing, walking, or driving etc. The attention is diffused over a longer period, and demands much less cognitive energy.

Still, there are only so many hours in the day that you can stuff podcast listening into. There are only so many chores or so many walking or driving hours (with lockdown on, driving time is down, and that has impacted Spotify’s podcast hours). But, if there is a limit on your podcast consumption hours, in contrast there is no shortage to the number of good podcasts getting created there. Internet removed editorial gatekeepers, and what it is led to is an explosion of content – good, bad, whatever.  Now, unlike text, it is very hard to figure how good the podcast content is. You can’t scan it like you can do with text, and figure out if it is worth your team. You have to listen for like 7-8 mins and figure if you are getting somewhere. Transcripts help though, and increasingly in the interview type of podcasts (there are several podcast formats I learnt), we are beginning to see some of the popular ones publish transcripts. Transcribing and publishing the transcript is primarily led by SEO and to help in discoverability but it helps time-starved people such as me consume the podcast faster. In fact if all of the podcasts I try to listen to (e.g., 20 Minute VC, Venture Stories), published their transcripts, I think I could consume far more podcasts and learn much much more.

Lifestyle startup opportunities

If we take this theme of making it easy for people to consume podcasts / learn from them, I see several opportunities for startups. These aren’t necessarily venture-fundable startup opportunities but more like lifestyle startups or side hustles.

– Offer a transcription-as-a-service modeled like a SAAS product to podcast publishers. Publishers can then put up the content for free (to aid in SEO) or sell it for a price (like Knowledge Podcast from Farnam Street).

– Go one step further and help publishers set up a storefront on their site, where their fans can purchase the episode transcript for a small fee. Alternately you can set up a common storefront, which can morph into a discovery platform as well

– Create a newsletter distilling podcasts – like a Blinkist for podcasts – with a short summary and your take on it. Super easy way to build a brand or even create a paid Substack. (At my employer Blume Ventures, we evaluated launching our podcast and felt the best podcast product was one summarising the best startup podcasts on the Indian scene. We launched it as The Takeaway.

The broad theme of the above 3 ideas, and there are probably a few more there, is improving the signal-noise ratio of insights or information, by reducing your consumption time, so you don’t have to spend 40mins to glean 2 interesting bits. This is done by giving you text to scan or distilling the key bits – effectively, converting audio to visual (text).

Audiles and Visiles

Barton Biggs, an investor and writer, in his book, Hedgehogging, spoke of the two types of learners

– Audiles, who learn primarily from talking to people, and who prefer that to reading, as a learning strategy. Their ability to absorb information from conversations or meetings is superior to their ability to absorb from reading. Here is Robert Steel, a Goldman Sachs partner, on Gary Cohn, co-COO at Goldman who later joined the Trump administration: “He’s never read a five-page memo in his life, but when he asks you to describe something to him he pays incredible attention and remembers every word.” Gary Cohn is clearly an audile. My sense is that people who like to listen to a lot of podcasts are primarily audiles.

– Visiles, who learn primarily from reading or watching. They may well enjoy talking to people but it is not a first preference. They are ok to do desktop research or reading as a learning strategy. Their information absorption ability from meetings or conversations is far lower than that of audiles.

And that brings us to the name of this piece, ‘Podcasting for Visiles’. If you are a visile, like me, then podcasts are always going to be a challenge to consume. For a long time I put off consuming podcasts, or would only consume it via transcripts. But there are many fascinating podcasts featuring intriguing guests that stayed hidden for me because of this attitude. There is a lot of interesting, super-useful content getting created via podcasting. A lot of it stays as hidden audio files; and while some of it comes out in transcripts, resurfacing via google search etc., it is far far less than what it deserves to be.

One way for Podcasting to up its usage game is to market itself better to visiles. I shared some ideas above around 3 opportunities in transcribing / distilling podcasts above. There are several more that are likely to emerge if you sit down and systematically think how you could ensure higher consumption of podcasts by visiles. Here is another below.

One big low-hanging fruit I thought was for Penguin or any publisher to create a best of 2019 or 2020 podcasts and publish it as a book. This could be themed. Like the Best of Startup Podcasts, Best of Gardening Podcasts etc. Sort of like ‘Now, That’s What I Call Music’ – a ‘best of’ compilation series that was popular when music was sold through CDs and not via streaming as is now. Or what if every leading podcast publisher had a book compiling the best content from their podcasts that they could sell; now it helps if the podcasts aren’t very topical though. Fans could buy the book to show their support, and this would also help the ideas get currency. (The same is true of many bloggers as well – I would love to read a book compiling all of Paul Graham’s or Eugene Wei’s essays, for instance).

What if you were a smaller book publisher looking to compete with Penguin / Random House and went and pitched yourself to Spotify or Barstool Sports and stuck a deal with them, to be their official text publisher and to help monetize the written word of their audio recordings? Perhaps there is a startup-idea (though I dont know if it venture-fundable) waiting here to be picked up. Books curated from the best of the Interweb – text or audio converted to text.

Back to the create-consume ratio

All of these ideas above are really ideas around reducing consume time – through text, distillation and enabling better discovery. So that brings us back to where we began, with the create-consume ratio. There is, and this is my opinion, an ideal create-consume ratio; imagine this as equivalent to the gross margin, or effectively the material margin of the brand, i.e., the realization from the product less the cost of materials incurred (no advertising or logistics cost here). This is the contract you make as a producer with the customer and establishes the price-value equation (credit to Sudhir Sitapati, author, The CEO Factory, for this insight).

IMHO there is an optimal contract for content creations, that is, an optimal create-consume ratio. And in my view, it is closer to that of writing an essay than of creating a podcast. Such a ratio, say 1:15 or 1:20 probably means a more refined, thought through product with a consistent promise to the customer.  I remember learning how Sonal Choksi and the A16Z team edits the podcasts a fair bit to create the final product – they optimize for insights per minute. No surprises that their podcast product is so good. Perhaps podcast creators need to edit down the podcast to key bits – speed down certain boring parts and speed up faster parts, relentlessly edit out the hmms and aahs, maybe instead of sharing the full podcast, only share key snippets with a comment or two. (Readers wont mind this, as many of them are podfasting anyway, i.e., listening to podcasts at 1.5x speed). This means increased creator time in doing these edits. Similarly transcribing, distilling – all of the above aimed at reducing consume time, would also mean increased creator time, either done by the creator or outsourced to a translator. 

And that perhaps brings us to the fact that all of the above is really attempts at resetting the create-consume ratio for a podcast. By reducing consume time through increased create time, thereby leading to a better product, and a better contract with the customer.

So that brings to an intriguing notion. What if we have got these interview podcasts all wrong? Maybe the podcast as is today is an intermediate product – sort of an unpeeled banana – and maybe the actual product is a distilled / edited podcast that has more creator hours behind it enabling easier or faster consumption. Maybe a few years from now, uploading the interview as is done now (with minor edits) will be the equivalent of you sharing your rough draft with your readers!

Podcasting for visiles

I once wrote “It is never been harder to create a podcast. It is never been harder to consume one”. We removed all of the frictions to create, and the result is that the chokepoint has moved to consumption. That presents challenges for us as users yes, but it also presents newer opportunities for builders, startups and podcast publishers to come up with new tools and products to ease the consumption load – through giving users the freedom to consume as text, or by curating and distilling better to save time etc.

So, who’s building?