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.


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.


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…

Modern Liberalism Is Not Peace, It's Pacification

Readers of my work over the past few years will know that I have long been interested in how natural human rebelliousness gets pacified (1, 2, 3).

I recently had the pleasure of working on this question with a group of co-authors, from very different methodological backgrounds. The final result has now been published in International Studies Quarterly. In "Liberal Pacification and the Phenomenology of Violence," (Baron, Havercroft, Kamola, Koomen, Murphy, Prichard 2019), we substantiate the concept of pacification relative to political science literatures on violence. Our real target was the popular conception of the "Liberal Peace" (i.e., modern liberalism causes peace, à la Steven Pinker).

While the article does not offer an empirical demonstration or test any hypothesis, I believe that — for a so-called "critical" paper — we went much farther than usual to develop at least a positivist case for our perspective. We do not pretend to have empirically defeated the "Liberal Peace" story, but we have planted a flag of sorts, from which "critical" perspectives might proceed in this direction with greater positivist/empirical sophistication.

Here's a key slice of the abstract:

We argue that the spread of liberal institutions does not necessarily decrease violence but instead transforms it. Our phenomenological analysis captures empirical trends in human domination and suffering that liberal peace theories cannot account for. It reveals how a decline in direct violence may coincide with the transformation of violence in ways that are concealed, monopolized, and structured into the liberal order. We call this process liberal pacification.

And here's a snippet of our positivist gloss that — I think — makes this paper stand out from a lot of other so-called "critical" papers:

…by reinterpreting the liberal peace as liberal pacification we are able to grant the empirical findings of liberal peace theorists while maintaining that the Pax Americana represents an intensification of violence overall. In the language of positivist social science, our theory is observationally equivalent to that of liberal peace theory. We expect that the quantity of direct violence inversely associates with the degree of pacification in a society. Therefore, our interpretation challenges research that identifies liberal institutions as the cause of declining violence. Liberal institutions, as apparatuses of liberal pacification, ensure that direct violence is increasingly rare while leaving the structures of violence and domination in place. The observational equivalence on particular dependent variables (in our case, all forms of direct violence) produces a theoretical change requiring the generation of novel observable implications (King, Keohane, and Verba 1994, 30).

In other words, empirical social scientists interested in the Liberal Peace should not toss this one in the bin labeled "purely theoretical postmodern crap I don't need to deal with."

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.