I’m going to give you a winning formula for founding a successful publishing startup: go back in time and start it in 2008. If for some reason that’s not possible, it’s not going to be easy.

Until recently there was no mainstream commercial ebook ecosystem, a curious state of affairs given that computers are very effective at moving words around. A unique confluence of gadget lust, display technology, and corporate willpower created an industry out of thin air. Before 2008, nobody bought ebooks. After the Kindle, lots of people did.

Any greenfield industry offers opportunities for entrepreneurs. Compared to clean tech, speech recognition, or high-def video (to name a few random innovations), digital publishing technology is trivial. Ebooks are tiny digital files, mostly composed in HTML. People buy them from websites that accept credit cards. This is some of the most ubiquitously understood tech today.

In early 2008 I founded a digital publishing startup called Threepress. Despite being a complete nobody, I was able to establish a viable business. I’m not being modest when I say this was largely due to good timing: few other technology companies were paying attention to publishing then; the bar was incredibly low for software quality and quantity; and there was a nascent community around a digital publishing conference and blog that I co-opted as my sales and marketing channel.

That conference folded in 2013, and the easy problems I could exploit are either solved or irrelevant. I’ll be fine; I sold my company last year. But how can a publishing startup succeed today, and at what?


There are a few things I think I did right that still apply in today’s climate, so I’ll start with those.

Don’t be a jerk

A good policy for humanity, of course, but it’s especially important in a small pond industry. Publishing is consolidating. At some point you’re going to work with — or for — everyone you’ve ever met professionally.

I’d go further than just saying “avoid being unpleasant” and suggest that you be generous. Every consultant has gritted their teeth over requests to “pick your brain over coffee,” and I’m not endorsing that behavior. But not all freebies are lost sales, and sharing best practices and tips on my company’s blog was by far the most effective marketing technique I ever used.

Don’t work for jerks

This one’s harder and therefore more important. We all have to eat, but I never regretted walking away from opportunities where I knew the client was just looking for a puppy to kick. Remember that if your customer is a jerk, you’re asking your team to work with a jerk too. There are plenty of non-jerk opportunities out there. Plus, jerks always try to get out of paying for your time. Which brings me to:

Always be billing

This is not what they will teach you in Startup School. They will sneer that you are running a “lifestyle business” and remind you about Tumblr. This is an industry that has invented multiple euphemisms for failure. You know what’s better than failing? Not failing.

Making money and building an innovative product are not incompatible, though there are real tradeoffs in opportunity cost. But publishers are risk-averse and have no sense of the timescale in which startups operate. A big Midtown publisher is not going all-in on your product, and if you’re a little guy negotiating with a major, you’re probably not talking to a person who can write big checks. (How many publishing startups do you know who say that they’ve got “the big six” signed up? They are lying — and an NDA doesn’t count.)

Spec work and delays are the deaths of publishing startups. Get publishers to fund something — anything. Low-level but forward-thinking employees can often scratch together modest amounts for pilots. If you can get them to pay money you are also guaranteeing that they’ll pay attention. You need both to succeed.


Here’s what I didn’t do, and where I think the opportunities are: really hard problems. These are the topics I was often asked to consult on, and while I believe I provided valuable answers, they were always long-winded workarounds and excuses. If you can actually do this stuff, you have a publishing business you can be proud of.

High-fidelity automated conversion of paged media to reflowable media

Or, “take any Word or PDF file and convert it to EPUB.” Currently this problem is hand-waved away using cheap labor. I think it’s literally unsolvable, which makes it a good candidate for someone to solve. I think it involves, like, machine learning or something.

Portable, crossplatform identifiers

Or, “iTunes Match” for ebooks. I read a book on my Kindle. You read the same book on your Nook. There is no streamlined way for an external service to know they’re the same book. That basic step is a prerequisite for all kinds of seemingly obvious applications, like shared annotations. The reasons behind this clusterblock are as much about business processes and vendor lock-in than technology, but it stifles innovation. (DRM in general is pretty consistently a startup-killer. While you’re inventing the future please also take this embarrassment out back and shoot it.)

Beautiful single-master book production

Or, “publishing.” I could make a sexy coffee-table print book and an attractive, accessible, cross-platform ebook, and instantly output revisions to both with a single click. But it would require hundreds of hours of meticulously coded HTML and CSS3, use costly proprietary software (connected to a toolchain I’d also have to invent) and be largely non-scalable. There are a handful of pretty good digital-first publishing solutions now, but their print support is limited relative to what book designers expect. (And the less said about the EPUB output from major print layout software the better.) Threepress never produced an authoring system because I was scared of this problem.

So, yeah. Lots to be done. Seems pretty hard. Fewer platforms for showcasing your work (at least there’s still one good publishing conference). I’m optimistic, though. I can’t wait to see what the next wave is like.