fbpx
How I made $3,300 on a short niche philosophy book

The Based Deleuze project is now officially complete, so it’s time to do some final accounting. Here I will review the financials, the labor/time costs, and the main lessons learned.

Based Deleuze was conceived and executed as a hard test. In research design, a hard test is a study that’s unlikely to find evidence for a hypothesis. If you use a hard test and you still get the results predicted by your hypothesis, then you can be extra confident in your hypothesis. As soon as I quit academia, my top priority was to generate reliable data about how much I could earn for my research and teaching on the open market. So for my first book, I strategically chose to do something as fringe/weird/unmarketable as possible, as quickly as possible. If I could get half-decent results, then I could confidently make future plans based on that data, because it’s very likely future books will do at least as well, and probably much better.

No matter how much I planned and strategized, I knew that my first attempt at a whole product cycle would be riddled with imperfections, so I purposely chose to do something that felt fun, light-hearted, and low-stakes, so I could move as quickly as possible.

I did so many things sub-optimally that I’m dumping most of those observations into a separate document, which I’ll post later. In this post, I’ll outline some of the biggest mistakes I made and highlight a few of the main things I did well.

Pre-launch

It all started on June 20th, when I tweeted an idea for a short book. It only got 6 retweets, but that was enough to take the idea seriously.

I made a pre-order product on Gumroad priced at only 5 bucks, drafted a quick cover on Canva, and then I literally DMed the link to everyone who retweeted, liked, or replied favorably to my tweet. This worked well and honestly it was a pretty great tactic for securing some initial buzz. I told them if it doesn’t get to at least 50 sales, I’m not doing it (this also gave them reason to share it, if they really wanted the book to happen). It crossed 50 sales so I committed to doing it. I set the release date to September 20th.

Then I got to work writing, which was my main project for about 2 months. I probably did about a thousand words per day, 2-3 days per week, on average. Pretty easy-going, to be honest, especially because I was free to do it however I pleased. I crossed my minimum target of 20k words after about 2 months. Then I did editing, formatting, and logistics in the time that remained.

The writing itself only took about 70 hours (measured hours of focused time actually writing, not a vague estimate of my time at the desk). See my detailed time-tracking below.

It was good that I announced a release date and a minimum word count from the beginning. The release date forced me to be done at a certain point, whether I was satisfied with the book or not (you never are). And the minimum word count gave me and pre-sale buyers at least some kind of objective standard for what would be enough. That was the only cold, hard promise I made about what, exactly, I would deliver on September 20. So I had at least some measurable standards for what I needed to achieve, and by when.

While writing the book, I tried to regularly tweet interesting and insightful stuff about Deleuze. I also made some Deleuze videos and uploaded them to Youtube. When uploading content I would generally link back to the pre-sale web page on Gumroad. I am pretty sure that work was effective at driving some sales but I did not measure any of that very carefully. And I had no systematic plan or schedule for this “content strategy.” I just did what I felt like doing.

Gumroad before Amazon

I decided to publish the ebook first, via Gumroad, and only much later publish to Amazon. I made this decision because Gumroad allows me to stay in touch with readers, whereas Amazon doesn’t. For obvious reasons, this is quite valuable for someone who plans to write many more books.

The audiobook and video course supplements

When I published the ebook on Gumroad, I also created and published a DIY audiobook and a 6-lecture video course. I learned this from Nathan Barry’s book Authority. One takeaway from that book is you should always have a few options, and one should be relatively quite expensive. This is because some small fraction of your audience wants everything you can possibly offer, some fraction is relatively wealthy, and some fraction just wants to give you more money because they like what you represent.

Gumroad let’s you create tiered products through what they call product “variants.”

So initially the price for the ebook was $5, I asked $10 for the ebook+audiobook, and $50 for the ebook+audiobook+course. These were bad prices. One huge mistake I made was under-pricing all of these things (more on that in a later post). I just lacked confidence for my first attempt, so I sold myself short. Maybe that’s necessary at first, though. Now that I’ve delivered on my first serious offering to seemingly happy readers, next time I’ll feel comfortable asking for a bit more. In the case of Based Deleuze, I would later bump up these initial prices, as you’ll see on the product page now, but only after 90% of the sales already came through.

One of the other big missed opportunities was not including the audiobook and the video course options as variants in the initial pre-sale product. I only added them in time for the Gumroad release date. I’ll definitely do all of that up front, next time.

The audiobook took some time but it was pretty simple. I just recorded myself reading the book. I did some basic editing but not much. It’s not quite Audible-quality but it’s really quite good, I think. My sales data below show that this was worth the labor. It also came in handy to have extra audio content. I posted the Preface of the audiobook as a podcast, for instance, to help promote the book.

For me, offering some kind of video course was a no-brainer because, as an academic, I can fire off lectures quite easily. But when I published the ebook on Gumroad, I hadn’t yet prepared any course content. So I just created a separate variant of the product, posted a planned curriculum of videos, slapped a $50 price tag on it, and in the description I said buyers would get the content over time after purchasing. I followed through with 6 one-hour video lectures uploaded over the course of a few months.

So let’s review the results separately for Gumroad and Amazon.

First launch on Gumroad, September 2019

I didn’t do a very sophisticated launch. I just uploaded to Gumroad, clicked “publish” or whatever, tweeted a bit, and emailed my list. At the time I had 1,215 subscribers. 52.2% opened the email. And 23.3% clicked the link to Based Deleuze. Here is the email I sent.

I earned $1,243 in the first month on Gumroad, as you can see in the graph below. By the time I was ready to publish, I had accumulated a healthy number of pre-orders, and then some publication buzz brought a bunch of new buyers.

Revenue From Based Deleuze: Gumroad
Revenue From Based Deleuze: Gumroad

The second spike in March 2020 coincides with the paperback release party in Los Angeles. Interestingly, launching the paperback on Amazon increased sales on Gumroad as well.

We can break down the number of sales for each variant of the product. As you see below, I only sold 11 courses but this generated more revenue than the 49 audiobooks.

Breakdown by variant: Ebook, audiobook, and course

Second launch on Amazon (paperback and Kindle), February 2020

The launch of the paperback was even more haphazard. The release party was at the very end of February but, to this day, I never really did a proper online launch for the paperback. I tweeted some stuff and mentioned it in my weekly newsletter, but there are a lot of things I just never did. For instance, I never even emailed the buyers of the ebook to let them know the paperback is available. And I never made a concerted effort to encourage Amazon reviews. I later learned that reviews are quite important for a few different reasons. (If you want to leave a review, I’d be grateful!)

Revenue From Based Deleuze: Amazon KDP
Revenue From Based Deleuze: Amazon KDP

Naturally, sales decrease over time, but I’m actually quite pleased with the lower numbers in the quiet months. In those months, I pretty much did zero work on promotion. If Based Deleuze continues to earn $100/month over the next several years, the financial success of this book would will be substantially more impressive. Maybe I’ll report back again later!

How much time did it take?

From beginning to end, I clocked 195.28 hours. These are focused hours, and I am pretty hard on myself about subtracting for distractions. I also don’t time all the little tasks that sometimes pop up randomly, so this estimate is a lower bound.

As you can see from my Toggl data below, writing the book and producing the lectures were the two most time-consuming parts of the project. Then, learning how to format the book for Amazon KDP was the third most time-consuming task. Fortunately, I learned a lot about how to do these things efficiently, so future projects should be significantly easier.

Time Spent on Based Deleuze
Time Spent on Based Deleuze

One big lesson here is that I should have outsourced more. Next time I will definitely not transcribe the lectures manually. That was stupid. My intern Ben Williamson helped with editing the videos, and my wife gave the final book a one-over for spelling and grammar mistakes. I think the grammar and spelling is quite solid; there are 2 or 3 sentences I cringed at after revisiting the book, but what can you do? As for the formatting and cover design, they are as good as my amateur design skills were ever going to get them.

The other lesson is that I definitely could have sequenced things to derive more positive externalities. I have a lot of ideas on this. Tweeting in a way that feeds the book content, writing the book content in a way that functions as lecture material, and so on. I’ve noticed many little ways one can structure and sequence a project like this to increase a bunch of little efficiencies, which might multiply quite powerfully. I’ll try to put them into practice for my next book and I’ll be sure to report back again.

Conclusion

At the time of this writing, 8 months after publication, the Based Deleuze project has netted a grand total of $3,290. That’s net revenue after platform fees, but before taxes and excluding my monthly fixed operating costs.

In terms of units, I have sold:

  • 452 copies of the book (ebook and paperback combined)
  • 49 copies of the audiobook
  • 11 copies of the video course

For the 195 hours I spent, I effectively earned about $17/hour so far. But if Based Deleuze continues to earn about $100/month for, say, another 3 years, that would roughly double the total revenue to $6,890 for an hourly wage of $35/hour. Still nowhere close to what a PhD generally commands, but as I said at the beginning, this was a hard test: Writing a weird super-niche philosophy book—which promises the reader nothing economically valuable—is one of the hardest possible ways to make money on the internet. I can certainly choose to do more lucrative projects, if necessary.

This is just the beginning. It’s hard to know how dramatically these numbers might improve as my audience increases, as I build up a catalogue of books and courses, and as my systems improve with every iteration. Personally, I’m pleased enough with the results to feel quite confident that writing and publishing books will continue to play a major role in my post-academic intellectual business model. I’m now most excited to observe the delta between book one and book two…

Lessons from Nietzsche’s Awful Publishing Results

“…the marketbell—The Daily Press—is only run for their own clique and not for the proud and solitary One.”

Nietzsche’s publisher Schmeitzner on the complete failure of Thus Spake Zarathustra

Independent intellectuals today should study closely one of the most profound and impactful thinkers in all of modern philosophy: Friedrich Nietzsche. I’m not referring to his ideas (although one should study those, too). I’m referring to the difficulties he faced publishing his books, and their utter failure in his lifetime. All the historical facts and figures below are drawn from Schaberg’s The Nietzsche Canon: A Publication History and Bibliography.

If you’re an independent intellectual, a review of Nietzsche’s publication history should fill you with a profound sense of gratitude and potency.

To start, consider Nietzsche’s book Human, All Too Human. When it was first published, one thousand copies were printed and only 120 were sold in its first year (1879). And that’s with the benefit of public controversy around Nietzsche’s break from Wagner and the scandal of being banned in Russia! Imagine today publishing a truly brilliant and original book, having a public and talked-about controversy with a famous and influential philosopher such as Slavoj Zizek, and only selling 120 copies! It’s unthinkable. Today, if there is anyone at all talking about your book, you will sell 120 copies at the very least. There are many reasons for this, most of them now banal (speed of information transmission, density of social networks, etc.). The comparison, however, is profound.

Next, consider that the reception of Nietzsche’s books got worse over time, which is the opposite of what happens to contemporary indie authors if their systems are set-up even 50% correctly. Nietzsche’s first book, Birth of Tragedy, made a splash: It received a polarizing but lively reception and sales were presumably healthy (I could not find quantitative sales data for that book). By the time of Zarathustra at the end of Nietzsche’s publishing career, however, Schmeitzner would write in a letter: “There is no question that the distribution of your books is getting worse.” Schaberg reports that Zarathustra was never acknowledged by “the press, the public, or [Nietzsche’s] peers.”

Thankfully, if you’re a blogger or indie book author today, it is nearly impossible for the sad fate of Nietzsche’s late works to befall your late works, unless you have zero systems in place. When you sell an indie book on the internet, you don’t just receive a bit of cash; if you sell through a platform like Gumroad, you gain a personal contact, an email address. And if you’re an open, generous person sincerely interested in your readers, many contacts even become genuine personal relationships. For these reasons, every new book by an indie author should sell at least as many copies as the previous work, and typically more. This, by the way, is why I published Based Deleuze on Gumroad first, and only published on Amazon after Gumroad sales plateaued.

But maybe Nietzsche’s small fanbase was super passionate, you think to yourself. Continuing with Human, All Too Human as an example, Schaberg documents precisely four instances of positive feedback. Two of them were personal friends of Nietzsche (Rée and Gast) and one was a lady he flirted with at the Bayreuth Festival. The only legitimate positive feedback from an objective and significant third party was from Jacob Burckhardt (most famous for his study of The Renaissance). Notably, Burckhardt called Nietzsche’s book a “sovereign book,” which would “increase the amount of independence in the world.” Personally, I think that’s wonderful praise, but even this is a backhanded compliment! He’s not saying it’s good; “independence” or “sovereign” is a praiseful way of calling Nietzsche bonkers.

On this point, the lesson is that you should prepare for nobody to care about your book, except your friends. Consider yourself blessed if you encounter even one polite negging from one smart and disinterested reviewer. Of course, you may very well enjoy more of a splash, I'm just saying you expect and prepare for... crickets.

The first year of Nietzsche’s Observations book saw “200-250” copies sold, then about 30-50 copies per year. Schmeitzner refers to this publication glowingly, suggesting that it was probably Nietzsche’s high-water mark. Based Deleuze has already beat Nietzsche’s high-water mark.

Nietzsche had to spend 881 marks of his own money to print 600 copies of Beyond Good and Evil. That’s somewhere vaguely in the ballpark of $15k in today’s US dollars. He must have turned over in his grave when Amazon first introduced print-on-demand publishing. It is now utterly unremarkable to note that anyone can publish and sell thousands of books for an up-front cost of zero dollars. But compare yourself to Nietzsche to see things with a new light. If that doesn’t give you a real jolt of intellectual virility then nothing ever will. If Nietzsche could follow through on more than 10 books, remind me again why you’re still struggling to publish one?

And then, all the little things.

To publish a book, someone like Nietzsche had to hand-write at least dozens of letters back and forth with his publisher, via snail mail. What a pain in the ass!

When there was an error in a published book—as there was with Human, All Too Human—someone had to go through all the printed books and fix the mistake with a pen, by hand. Today we just edit the file once and re-upload it to Amazon or Gumroad.

Nietzsche frequently dictated his writing, which means that another person was required to type as he spoke, often for about 2 or 3 hours every day for months at a time. Whether he or his publisher(s) paid for this labor isn’t clear. Regardless, we now benefit from computers, which can, for pennies, automatically transcribe spoken words at about 95% accuracy.

Don’t even try to tell me it’s difficult to write or publish a book, don’t even try! I will send you this blog post to shame you!

The Optimal Podcasting Process for Indie Thinkers: Automate and Proliferate

Here is the podcasting process I’ve developed over nearly 2 years of iteration. It’s not perfect, and it’s not for everyone, but for many indie thinkers I think this is the best system. Certainly better than any other piece of “how to run a podcast” content I’ve yet to encounter.

This is intended to be directly useful to at least some of you, but I’d also love to hear questions, comments, critiques, or specific requests for additional explanation/instruction. I will probably build this out in some way.

I just listened to the a16z meta-podcast on how they run their podcast. I was struck by how — though filled with strategic insight — their entire framework and all of their advice was totally useless for individual thinkers/makers/creators bootstrapping a podcast as one part of a socio-technical production system. The main reason their strategic framework is useless for people like us is that it assumes a huge production budget! If you have the money to pay a team of editors, their advice seems great. But what if you have no budget at all, and hardly enough time to produce one podcast every week? And not to mention, if you’re interested in obscure intellectual stuff that doesn’t lure anyone with dreams of great wealth, like a startup podcast does?

That’s the strategic challenge I’ve been trying to solve for the past 2 years. After so much iteration, it’s about time I share the framework I’ve developed. Then I’ll explain in detail the tools and workflows I use to run my podcast in line with this framework.

The framework: Automate+Proliferate

In podcast system design, there is a tradeoff between quality and quantity. Within a given period of time, a fixed supply of labor power, the more effort you allocate to improving quality, the fewer podcasts you can create and publish.

The main hypothesis at the core of my system is that creative thinkers and makers with modest audiences and little funding should heavily favor quantity over quality. Given how many thinkers and makers I’ve met who seem very concerned about their podcasts’ quality — and, at the same time, fail to deliver volume consistently — I’ve come to realize that my system is very far from obvious. So let me explain the rationale and give you the concrete details of my own system.

The diminishing marginal returns of audio quality

Assume that audio production quality (including audible features but also substantive content density, which is an editing artifact) can be understood as a distribution, such that an unedited recording of my marijuana-hazed ramblings with an analogue tape recorder in a busy nightclub produces a podcast in the zero percentile of production quality. Basic consumer technology is now good enough that nearly any podcast recorded by anyone, in any empty room of their house, with any digital technology lying around, with no editing, would already be somewhere around the 60th percentile.

Professional podcast producers differentiate themselves by moving from the 80th-90th percentiles of production quality to the 95th-99th. This makes sense if you already have money and a substantial audience to start with. But you have to understand diminishing marginal returns. The value derived by moving from the 80th percentile of production quality to the 95th is much less than moving from the lowest percentile to the 60th percentile. This is one of the key facts undergirding the logic of my system.

My theory here is mostly based on my observation of other projects and my own trials and errors iterating my own system. The only data I have to support my theory come from my own results. My podcast is not huge but for a solo podcast about my own fringe ideas and random friends on the internet, with no clear branding or particular value proposition or even a coherent tagline, it punches far above what you would predict by listening to its audio quality. (~80th percentile globally, or more than 1000 downloads per episode on average.) Regardless, my theory is speculative and I could be wrong, so take or leave my suggestions as you see fit.

General principles to Automate+Proliferate

As much as possible, develop a technology stack to optimize for volume, consistency, and quantity of outputs, but merely satisfice for everything else (doing the best you can with the least amount of effort). When it comes to preparation, editing, and promotion, first automate as much as possible; then, whatever can’t be automated, reduce to the easiest possible heuristics and decision rules. I keep the latter on index cards.

The basic rationale is that you don’t have the time or money to start with really high quality. If you follow the workflows of professional podcasters, you’ll bankrupt yourself temporally and financially before you ever get off the ground. On the other hand, quantity can be used to compensate for the quality you can’t afford. At least in the short run. If you follow my framework to produce one episode every week for 100 weeks, you might have the results of someone who produced one really high-quality episode every week for 50 weeks. If you can get your automations humming really nicely, maybe you can achieve the same result with 2 episodes per week over 50 weeks. To be clear, these particular numbers are arbitrary, I’m just using them to illustrate the idea.

A few people will eventually complain about the production quality here and there, but it’s surprisingly advantageous to have a few things for people to complain about! Think of it as a strategically placed tripwire, which alerts you when you have your first listener who cares… Also, something I learned in practice (I promise I’m not clever enough to have strategized this in advance): The low production quality will become a natural and reasonable hook if and when you decide to test the waters of patronage. “You want better quality? Here’s how you can help me deliver it!” In one way or another, as you gain an audience, you can gradually invest more in quality.

By this point, you might be wondering how I’ve setup my tools and workflows to Automate+Proliferate. Here’s my system, concretely.

Tools, sequences, and workflows

TLDR:

  • Livestream podcast on Youtube
  • Download audio track of the livestream
  • Automatic editing and uploading with Auphonic
  • Automated syndications via Libsyn
  • Automated distribution with Zapier

Recording on Youtube

First, by recording my podcasts via Youtube livestream, and then posting the audio to the normal podcast feed, I build my audience on two platforms with no extra labor. It limits my editing options, but really it justifies my podcast’s lack of editing and manages my audience’s expectation. You have to be creative piecing together such non-obvious complementarities, to make a system that works for you and your brand. My podcast listeners get that I record first on Youtube, because I tell them, so the imperfect audio is not as upsetting as it might be. “It’s not that I’m lazy, it’s just that my podcast is recorded live so editing is not an option!” Also, people who listen to stuff on Youtube and people who listen to stuff on their podcast app — these are totally different people, I’ve come to believe. So I’m pretty confident this system-design choice really has built me two audiences for the price of one, rather than just splitting one audience into two locations.

After a livestream is done, I download the audio using an app called Clipgrab. (None of the online, in-browser utilities can robustly handle long videos, I’ve found).

Editing and posting with Auphonic

Then I upload the audio to Auphonic, which is an automated podcast editing service (paid subscription). It doesn’t make my Youtube audio sound like Serial, but it ensures that the speakers are roughly similar in volume, that all podcasts are set to one common volume level (industry standard -16db), and that any particularly bad background noises are dampened. The honest truth is I am not certain that Auphonic has delivered a tangibly improved listening experience, as I never noticed a major uptick in positive feedback when I started using it, but I think it does! At the very least, it eases my conscience knowing that I’m doing my honest best to deliver listenable audio, within my constraints. But an equally valuable aspect of Auphonic is that it helps you automate a whole bunch of other tasks in the podcast-uploading process. This is why it’s definitely worth the price. You can upload intros and outros and have Auphonic automatically add them to the edited audio track before it’s uploaded to your podcast host (also done by Auphonic). You can upload templates for your podcast show notes. You can export to multiple destinations, all at once. You can also generate a bunch of formats, including transcripts via AWS or Google Cloud. (Though I stopped doing these because I wager that the machine learning will be way better one or two years from now; I’ll transcribe all my podcasts then!).

You can have Auphonic publish immediately on Libsyn, or you can just have it push to Libsyn — and you can publish or schedule it later. I do the latter. I also have it push to my Google Drive.

Aside on patron delivery

Once it gets edited and pushed by Auphonic, I have Zapier automations that will push it to a private RSS feed reserved just for my patrons. That seems to work robustly and I was quite pleased with my time-saving cleverness, except that I eventually started manually creating Patreon audio posts anyway. That way patrons get an automatic email (I want them to know I’m hustling, in case they neglected to subscribe to the private feed I gave them) and also it gives me a regular flow of Patreon-branded items to share on social media (I shill for patrons much less than average, so I have to do something; sharing a link to something I posted on Patreon seems like the least offensive way to run some minimal, recurring “promotion” for patronage).

Mostly automated publishing, syndicating, and promoting (Libsyn, WordPress, Overcast, Zapier)

After I post a quick audio post to Patreon, I will schedule its release on Libsyn. Within Libsyn, all new published podcasts are automatically syndicated as blog posts on my WordPress site. I have an automation that will post a tweet for every new blog post. I have automations that add all new blog posts into a particular text file on Dropbox, as styled HTML links, such that I can copy and paste each item into my Friday morning newsletters. No writing on my part is required; metadata from the WordPress post is arranged by the automation to provide written context for the link.

Once it goes live on Libsyn, if there are particularly good moments, I might use the Overcast app on iOS to create some clips and share them to Twitter. (If I’m busy, I tend to neglect this step. Not sure if this is wise or not.) This is a really nice and convenient functionality, and the clips get a lot of listens on Twitter generally. I am not sure to what degree this clip sharing drives podcast subscribers. Podcast subscriptions and downloads are metrics that I have not yet given much attention to, honestly, in the larger scheme of my system. Getting too concerned with these metrics and specific conversion rates would require me to start running proper statistical models, which — trust me — I look forward to doing. But at my current stage, it would be vanity (similar to optimizing for quality). There are still obvious, lower-hanging fruits for me to optimize, and I’m not yet big enough that these kinds of analyses would be worth the labor. Maybe soon though!

For some time, I would post podcasts to relevant subreddits on Reddit but I stopped doing that. In part, it felt kind of spammy and somewhat egotistical to post my own podcasts. A few of my podcasts were shared by other people to subreddits, so after I saw that, I think I stopped sharing myself because I was hoping my fanbase would eventually start doing that all the time. I don’t think this has really happened. Sad. I’m agnostic about whether posting to subreddits on your own behalf is worth it. At the very beginning it’s probably worth it, beggars can’t be choosers.

I also have a Zapier automation that pushes a link to the podcast’s blog post on my Discord server.

Operations management

I track all my podcasts, at each stage of this process, in Airtable with a Kanban view. So I can see the whole system as a pipeline. This helps me ensure that the flow of the whole system is on track. The stages in my table are:

  • Need to invite
  • Youtube done
  • Patreon posted
  • Libsyn scheduled

Conclusion

That’s pretty much it. Maximum bang (of volume, consistency, and distribution) for the buck (of my own time and effort). I’m able to do one per week (most of the time) and there have been weeks I’ve done two, even while doing lots of other stuff. My audience grows, modestly but effortlessly. As I said at the beginning, I’m now in the 80th percentile globally, according to the data I’ve seen. If you want to be a fancy professional podcaster, get some money and hire a professional audio engineer. If you don’t have money and just want to pursue disinterested intellectual interests, develop your ideas, make friends, and slowly build a modest but nonetheless real audience, consider using my framework. Let me know if you do!

Reasons not to start an online magazine

TLDR: The magazine model is not the best choice for indie intellectuals trying to start something. Two reasons: (1) Accelerating digital culture requires you to move fast and group projects move slower than solo projects, and (2) accelerating media segmentation rewards increasingly unique, hyper-niche content, so smaller, in-house production teams are better able to express unique micro-niches. I discuss some other social-psychological bases for this view, and acknowledge some exceptions.

I recently consulted for an individual interested in starting a group publication online, but he wanted my feedback. Here’s a transcript of my answer. The audio will eventually make its way into a podcast or something….

The publication model a lot of people have in their minds, I feel like it doesn't work anymore. There's this mental model people have of putting together a kind of collective internet project and... People will think about that in this relatively traditional template as a "publication," right? You have an editorial staff, and then you solicit talent and you commission works and stuff like that. My sense is that — I mean, I'm not 95% certain of this, but it's a hunch and it's my own personal, strategic sense of things — that type of model is just broken. It doesn't work any more in the new digital context.

I do see some people try to do digital publications, kind of like on the magazine model — I guess there are a few that are kind of succeeding right now. Something like Jacobin would be an example, or N+1, I think, is still going strong, or even Jacobite is something of an example. But I think what all of those examples have in common is relatively large, captive audiences they're drawing on. So Jacobin, for instance, it's pretty much a DSA organ, they've monopolized a particular membership group of people that are already unified, already see themselves in the same subculture and they're already dues-paying members. So they have a certain willingness to pay. They're piggy-backing on that, you know? N+1, they're much older I think, and they came up before the real digital publishing revolution really had its effects. Also, those people are really well networked in New York circles and stuff like that. Then with Jacobite, I don't even know if they're an exception that proves the rule because I think they're probably just barely staying afloat. They punch above their weight in terms of impact, I think they're a successful venture, but I don't think anyone is really making much money and I don't think it's particularly growing in a really rapid or exciting way.

In some sense, they're an example of: You can do that type of thing well, and they've definitely had a pretty good impact. Though they're pretty well networked in Silicon Valley circles, and I think they probably have a few decently generous backers, is my understanding. Of course, there's a bunch of more traditional magazines or magazine models that just seem to be going down the tubes slowly…

I could go on at length about why I think that model is a bit dangerous at the moment and why it's not exactly the most promising or exciting thing, but that's not to necessarily push you off it! These sorts of things can be very worthwhile, even if they're short lived or they don't exactly take off in the way you might hope. I'm not necessarily saying don't do it. I'm just saying, I have been watching a lot of this stuff and I do have a sense that that's not the most exciting thing.

So you might be wondering, okay Justin, what's replacing it or what's the better model? I think people are just figuring this out, but one way I would summarize it... All the new stuff that is most successful and that also small groups can bootstrap with relative efficiency, without connections or backers... It has to be more subcultural and niche. Bite the bullet and embrace that you're going to really only be interesting to a pretty specific type of person. That's part of what is going on in a lot of the successful ventures.

The second thing is, the projects that are succeeding right now, they really leverage the agility and efficiency of either solo creators or very small group collaborations that get along really well. One of the problems with the magazine model is that it's this larger corporate structure, where there's a board or there's some sort of leadership panel, and then they organize other people's work and fund it and support it and edit it and publish it. And one basic problem with all of that is, it just takes a lot of time. There's a lot of friction, and transaction costs. You'll find that trying to get people to write for you is like herding cats, you know, because most... er... a lot of people just aren't that productive. They're just not that focused. Even smart people, even people who are capable. It's just very hard for people. Whereas, if you're a solo creator or you have a small band of three to four people who are pumped and productive and motivated, and they're on the same page sufficiently well enough that there's not a lot of time wasted on debates about what should be done and what shouldn't be done... That small group can just churn out content. And I think what's really winning right now is volume and consistency.

An outfit that can produce new stuff every single day, let's say for a few months straight. Even if it's relatively short stuff, even if it's not the most amazing stuff. If you can achieve that kind of consistency and volume, I feel like that's the stuff that is winning right now. Just because there's so much noise that, to really get through to any particular community that you want to communicate with, there's such a high quantity of stuff floating around — and a lot of it is crap — that the only way to really get through is if you have your signal being emitted out into the world regularly all the time. Because only a few people are going to see each tweet. Only a few people are going to see each blog post. But if you're able to do that every single day, then it adds up in a way that other tempos or publishing frequencies don't add up.

Also, the smaller groups or the solo projects, what they have to their advantage right now is real personality character. This is something that people are really into right now, people won't really subscribe for things or buy things or even read things unless the producer is a human that they feel some sort of connection to. And again, the magazine model, it's much more corporate and cold. The magazine model prides itself on high quality, long-form stuff, which is cool, but it just doesn't register with people today, psychologically, because even if there's a really, really, really profound, good 10,000 word article, that was patiently edited, and the spelling is perfect, and the writers are very talented.... People see that and they're just like, "I don't feel like reading this." Whereas, if it's some person that they know of and have some sort of identification or relationship with, or a small collective that they know of, then they're emotionally motivated to check it out. [You’re not giving up on maximum high-quality, because all that content can be gathered, edited, revised, and reconstituted into sophisticated long-form later, just as high-quality as a magazine’s feature article.]

The emotional motivation is the make-or-break variable that is deciding right now what rises to the top and what doesn't. Especially for purchasing decisions. For people who are going to make that decision, "Oh, I want to throw these people five bucks a month," or "I want to buy this book," or not. That is hugely emotional and hugely personal [because trust in institutions is unprecedentedly low]. And I think you can only win that game if you have a very human, small-scale type of high-volume, high-consistency production model with a very focused subcultural niche, which is just very human and raw and honest (though the humans should use lots of AI in their production processes behind the scenes, as I do). That's my two cents on what seems to be working right now and what seems to not be working…

Fully automated personal brands

Right now, it’s still seen as bad taste to overly automate your personal social media — and for good reason. But taste changes, and it always follows the money.

As machine learning gets better, we will soon cross the threshold where some minority fraction of the dumbest people will be unable to distinguish between a real human's "personal brand" and a fully automated machinic substitute trained on that human's history of creative content. Let's call that fraction the "dupe fraction." In the first period after crossing this threshold, higher-IQ people will still be capable of such discernment, and they will mock and stigmatize anyone they catch replacing themselves with machinic substitutes. But as the dupe fraction increases — and it must, unless you think machine learning cannot get any better — the payoffs to machinic self-replacement will eventually outweigh the costs of stigmatization by elite discerners. It is inevitable that there will therefore be a period in which elite discerners will be barking into a void, only to be outcompeted (with respect to influence) by those who bear the short-term stigma to win the longer-term race of machinic content domination. Then, of course, machinic self-replacement will become the index of Cool.

The tricky problem is knowing when we cross this threshold. It is not inconceivable that we've already crossed it. Machine learning tools may already be good enough, for how many dumb people are already on the internet, that someone such as myself could hand over all my public posting channels to machine intelligence, turn the quantity and consistency up ten notches, alienate all my high-IQ audience, but replace them with 100x as many dumb people over the course of a couple years.

My personal diagnosis — and trust me, I've been looking into this for some time! — is that we're not quite there yet. I've even experimented with some pilot programs, e.g. an anonymous Twitter account trained on my own writings, for instance. It's pretty decent, actually, but if I ever used it for my personal account, the number of people for whom it would pass the Turing Test is too small relative to the number of smart people who would see through it and think I'm a dumb loser.

I should note that another crucial variable is the accessibility of machine intelligence. I could perhaps do better than one Twitter account trained on a collection of my own writings, but the currently available tools and workflows are still a little too demanding for this to be rational at the moment. Although the tools are rapidly growing more convenient.

It's ultimately an empirical question when, exactly, we cross this threshold. Everyone has to make their own wagers. But I think most people are over-estimating how long it will be until machinic self-replacement becomes the winning strategy — indeed, an existential necessity — for any intellectuals and content creators wishing to remain in the meme pool.

One thing is clear, however. Do not wait for machinic self-replacement to be affirmed by prestigious institutional opinion. By that time, it will certainly be too late: all the cool kids will have already machined most of their internet personas to unprecedented degrees. By then, it may already be the pre-requisite for making real and valuable social connections with smart and creative people in the real world. I would bet there are already Zoomers experimenting with automated "personal brands" to degrees I would look down upon. I'm guessing I won't hear about them until their content systems blow mine out of the water. The trick will be to make this transition late enough that you keep as many of your high-education/high-IQ audience as possible, but early enough that you win a decent slice of the first-mover advantage.

An Effort Allocator for Content Creators
Get the spreadsheet

If you're trying to build a long-term intellectual life on the internet, one of the biggest problems you'll face is figuring out how to allocate your effort. There is no template for doing this successfully — yet — and it's a doozy of an optimization problem. If you don't spend a lot of time producing original, high-value work, then obviously you'll never accomplish anything meaningful. But if you don't spend a decent chunk of effort building and improving your distribution systems, then your probability of having impact or making money is way lower. Other vexing questions include whether you should use result-based goals (e.g., "3 blog posts per week" or time-based goals (e.g., "3 hours of writing per day"); and what exactly should be your "north star metric." Audience size? Money? Subjective insight intensity?

Since "going pro" on the internet 8 months ago, I've been forced to tackle these problems head-on. I'm not sure how much my solutions will generalize, so I'm not even going to defend or promote them, but I figured I will at least start to share them.

Here's how I've solved the puzzle of effort allocation, for myself.

First of all, I decided that what I'm really optimizing for is having a good life or what the Greeks called eudaimonia. This means I need to make a decent bit of money, but I don't need tons — so I’m certainly not maximizing that. It means I need to spend most of my time working hard on what I am personally most called to do; in my case, that means focused, original, creative intellectual work. And it means I cannot do either of these things so intensely that it causes me to be unhealthy or neglect my most important relationships. If this whole adventure makes me more stressed and joyless than academia, then it would be a failure.

The way I've operationalized this perspective is to assign percentages to the various priorities of my intellectual work system (not to exceed 100%). Then I formalize how many hours per week I want to work in total, and the batching frequency I consider best for each task.

I then feed these personal decisions into a spreadsheet that converts them to blocks of time I must schedule on my calendar. The idea is to input numbers that reflect my ideal, properly ordered work life (percentages out of 100%) and output specific, concrete requirements. If you use this system, and you have the discipline to execute what is blocked out on your calendar, then you can be confident you're not doing too much or too little of the various tasks in your system. Whether you've assigned your priorities optimally is a whole separate question, which no spreadsheet can solve, but at least you can see what your priorities really look like in practice. And you can more easily tweak them iteratively, as necessary.

Feel free to copy mine and use it yourself. You decide on the values in yellow: How many days per week can you work on your intellectual system? How many hours in those days? And how much do you want to focus on original creative work vs. distribution, video vs. audio, etc. Then just block out your calendar according to the values in green.

The best immediate effect of using this tool is realizing how insanely unrealistic is your current mental picture of everything you vaguely hope to do over the next week, month, etc. It forces you to face the fact that you can't do everything you want to do, but at least you know that what you can do will amount to the best possible approximation of what you'd like to do.

1 2 3 4

The content of this website is licensed under a CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL LICENSE. The Privacy Policy can be found here. This site participates in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram