A story of the little guy in publishing

I thought that since book publishing has been a theme at Angry Bear with Mike’s book Presimetrics, and in light of other books from Angry Bear authors a possibility, the story seemed interesting. Also Angry Bear has quite a few readers who are also ‘engineers’ who also might be interested to hear about his story from a different perspective.

by Piaw Na

I am the author of An
Engineer’s Guide to Silicon Valley Startups
. As a self-published
author, I do everything from writing to cover design to shipping books
(yes, I do my own taxes as well), and I thought I would write
something about what a self-publishing business looks like.

I first thought of writing this book because of
having to repeat myself too many times to every new employee at Google
who asked me about exercising his stock options. Obviously, I never
got around to it until some time in late 2008, when I wrote a blog
post about startup
stock compensation

The blog post got a surprisingly anemic reception, yet it gets
continual hits even today, indicating that while few people piped up
and said, “Oh yeah, I’d like to read a book about this,” there were
always a few people searching for the answers that the blog post was
about. I fleshed out the chapter a bit and showed it to a couple of
friends of mine. The feedback was, “The title sucks. I didn’t think
I’d be interested in the topic, but once I got started I enjoyed it
despite myself!”

I put the project aside and went through 2009 busy with work and
travel. I went down to 80% time at my day job, but the first few
months of it was spent catching up on life activities, rather than
working on the book. I bought a house, which turned out to be a
massive project all by itself, scanned several year’s worth of
, which was something I should have done ages ago, and
before I knew it, had hit Thanksgiving with only 3 chapters of the
book done.

I ran into Kickstarter by
accident, and decided then and there that if anything would spur me
into writing the book, paying customers would! I put up my book there,
and like magic, started writing furiously. The book was written
entirely in OpenOffice, using styles and templates I had found on the
web as a guide to ordering my thoughts. When I found myself working on
the book even during my winter
, I knew I would get it done, and sooner than I expected.

I searched the web about publishing solutions after reading John
Reed’s book
. His approach of printing his own book and binding it
at home didn’t appeal to me, and neither did ordering a thousand copy
run of something I was sure would be a small market. I thought of
trying to sell it to O’Reilly books, but the thought of having to deal with a real publisher made me wince.
The last time I had a book contract, there was a lot of pressure to put in fluff to make the book fatter, because that’s how people buy books in bookstores. After some research, I found CreateSpace, and discovered that despite being owned and affiliated by Amazon.com,
you could just treat them as a short-run printing house and not let
Amazon sell it on your behalf. If Amazon did sell it on your behalf,
they would take 50% of the cut. If Amazon distributed it to bookstores
on your behalf, they would take 70% of the cut. Neither of those deals
sounded good to me: there are maybe 100,000 engineers in the country,
and at most 5% of those would be interested in startups at any given
time. That caps my sales at about 5,000 copies, if I reached every one
of them (I probably won’t).

Once I got it into my head to do it through CreateSpace, I stitched
together all the files and formatted it in various form factors to see
how it looked. I ruled out 8×10, because it’s bulky and hard to ship.
Smaller sizes required more pages, and CreateSpace charged for
printing by the page after 100 pages. I decided on 6×9 as a

Putting together the cover was interesting. Amazon provides a
template, and I had a copy of Photoshop
anyway from my photography
hobby. It turns out that if you need to do something with Photoshop
nowadays, all you need to do to google the task you want to do, and
follow the step-by-step instructions. I was surprised at how easy it

Once all that was put together, I signed up for CreateSpace’s Pro Plan (which
reduced the price per copy of the book: all it takes is a 40-copy run
and the Pro Plan pays for itself), and then worked on iterating on the
interior and exteriors. You have to do this a couple of times because
how much margin to use and how to make it look good isn’t obvious, and
of course print is always different than looking at photos on a
computer screen.

The Kindle version was very easy. I was very familiar with MobipocketCreator from prior experience, and it sucked in my files just fine. There were
a few glitches, which I dealt with by diving into the html intermediate format and directly editing the files (you can do this once the manuscript is in close to final stage). It turned out that by using the OpenOffice styles appropriately, my book lined up very easily with what Mobipocket Creator expected.

All the pre-production work took about 2-3 days of total work time. Proof reading and copy-editing was helped a lot because Larry Hosken took it upon himself to copy-edit the book in detail. Others provided gobs of input as well. The Kickstarter process is extremely valuable in this regards. You really do get people who are interested in helping out, and are familiar with the topic at hand. The final part was sending out the manuscript to everyone who was quoted in it to make sure I
didn’t misquote anyone. Everyone was incredibly helpful and I’m very grateful that people have been so generous with their time.

Each copy of the book from Amazon costs me $2.90. Add in the costs and risks of holding inventory, and a $5 premium for the paper copy didn’t sound unreasonable. I wanted to keep myself economically indifferent as far as selling electronic books or paper books, and so priced the paper book at $29.95 and the electronic version at $24.95. The big bugaboo about selling an electronic version is piracy, but I decided
that given my audience, I had to have an electronic version, and Software Engineers are infamously opposed to DRM, so I went DRM-free.

One of my biggest headaches turned out to be marketing the book. As an
ex-Googler, I first thought of Adwords, and bought some Adwords ads. It turns out that it’s incredibly hard to use the Adwords model for something as niche as a book. For instance, if you had a 1% conversion rate and you made $25 a book, the maximum bid you could make on Adwords would be $0.25! And at that price you simply would not make any money! As someone who maintained a blog and enjoys writing, however, what I’ve learned is that blog posts that garner wide interest amongst the target audience was extremely effective. For
instance, my blog post on Google Hiring Committees drew 10 sales the day it got syndicated to Hacker News. So ironically, the best way to get people to pay for books is to write a lot of free stuff, which as a writer I am more than willing to do.

Someone asked me whether I would have been better off putting the book on-line and selling Adsense ads right next to it. In the 4 months I have sold the book from my href=”http://books.piaw.net”>web-site, which is a href=”http://appengine.google.com”>Google AppEngine hosted page, I
have sold 108 copies at full price (as opposed to the 42 copies sold
at promotion price on Kickstarter). As of today, that’s a total
revenue of $4628, with a printing cost of $400 (including books not
sold, given away as review copies, etc). Adsense on my blog, in the
mean time made about $100. Amazon Associates (also on my blog) made
about $200. It’s quite clear that Adsense/Amazon Associates is not a
good way of monetizing the extremely niche content that comprises the
book. Stamps turned out to be the second biggest variable expense at
$201.55. Business overhead came in at around $1,000 for various
business licenses and California’s LLC tax, which you would not have
to pay if you were doing this as a side-job and didn’t have to
incorporate in order to qualify for group health insurance. It turns
out that books are a highly scalable business, so each additional book
you sell turns out to be pure profit.

The book sold out its initial 100 copy run at the end of April, 2
months after I started shipping books. With print on demand there’s
never any need to order huge runs, and I only have so much space at
home, so small runs on the order of 40-100 copies minimizes clutter at
home while always keeping a few copies on hand for sale. My current
sales range from 3-5 copies a week, which is approximately a run rate
of $3900/year. About one third of the sales are electronic, which
indicates that going DRM-free has not really cost me much revenue. I
initially thought that since the book’s main value was dealing with
taxes, which were incredibly California specific, I would not be able
to sell the book outside California. That turned out to be false. I
have sold electronic books to Sweden, Norway, England, and Russia.
Book sales to New York turned out to be surprisingly high as well.

All in all, I think while having an editor, proof reader,
distribution, would have been nice, I’m not sure I wanted to give away
90% of the income from the book to get that, given how niche a market
this book will sell into. That would have made the book only as
profitable as the Adsense blog. Ironically, if your book is going to
flop, you are better off going with a publisher, which will grant you
an advance of about $10,000, which is more than most books will earn
out in their life time. However, if the book has any success at all
and will turn out to be an ever-green title (one you can sell year
after year), you are better off going the self-publishing route. All
in all, I am pleased with the process, though given the 3 months of
real time I had to put into writing the book, I would have made more
money fliping burgers at minimum wage, after subtracting out all the
obligatory expenses.

I’ll come back in another 6 months to report on sales and discuss
anything else that I have learned in the mean time.