Andrew Chen Archives

Subscribe · Featured · Recent · The Cold Start Problem 📘
Dear readers, I have moved to Substack and I will be writing here from now on:
In the meantime, I will leave up for posterity. Enjoy!

How to write a business book — behind the scenes from THE COLD START PROBLEM

(above: Me in Sep 2023, a happy author, finding the Japanese translated version of my book at the wonderful Daikanyama Tsutaya Books in Tokyo)


Dear readers,

As many of you know, 2 years ago I published my first book THE COLD START PROBLEM. It aims to tell the story of why some products — YouTube, Instagram, Uber, Slack, Dropbox, and others — end up with hundreds of millions (and sometimes billions!) of users, and to provide the definitive theory of network effects which are often referenced in the tech industry, but only superficially understood. It’s been a success, now in a dozen markets, translated into many languages (including Japanese, Chinese, Spanish, Russian, etc).

Here’s a screenshot of some of the wonderful pictures that readers took during launch week:

This was an awesome experience. Nevertheless, I swear I will never write another book again . (I guess never say never, ha)

The creative process was a long, meandering path, and folks ping me from time to time because they want to take on a masochistic journey of their own. So this post will be about the messy, annoying, behind the scenes leading up to writing a book like this — it describes a bit the creative process, but also some of the major milestones and lessons learned.

Hopefully it will be useful for someone in the future who is thinking of a big writing project of their own.

A brief summary of what I’ll cover:

  • Month 0: At first, writing a book seems like a fun idea (until you figure out it’s not)
  • Month 1-6: Finding an agent, writing a proposal, and opening up your Christmas presents early
  • Month 6-12: Collecting and organizing the ideas — lots of fun chats, reconnecting with colleagues, talking to great people
  • Month 12: How to write the initial the outline, then the mega-outline — finding the formula
  • Month 12-24: The very messy middle, the trough of sorrow, the hard slog, followed by trench warfare (yes, it’s 3-5 years to write a book)
  • Month 24-36: Why you’ll feel insecure about the creative process
  • Final months: Just ship it already

I’m also going to link to various copies of intermediate content along the way — unfortunately I can’t share everything (like interview notes, etc) since some stuff will have to be confidential, but here’s a few interesting bits anyway

OK — so let’s get started on the journey.

Month 0: At first, writing a book seems like a fun idea (until you figure out it’s not)
I joined Andreessen Horowitz in mid-2018, I had already been writing on my blog for 10 years and I was kind of having some creative boredom over it. At that time, Elad Gil had just published his book and we had a nice convo at MKT’s lounge about a week after his book was out — he had amazing things to say about the process (he writes faster/better than me, in the back of Ubers it turns out), and he said it helped him a lot professionally. At a16z, as you all know, Ben and Scott have both written fantastic books as well, and it seemed to really be great for them professionally, so I thought it might be a fun challenge to do the same. So think of the motivation as 50% a creative challenge, and 50% seeing what it had done for other people.

I had two lines of thinking in terms of picking the book topic. First, I’ve had good luck taking ubiquitous jargon and writing the definitive blog post on the topic — something I did with growth hacking, CAC/LTV, viral loops, and concepts like that. I had a few ideas bouncing around in my head that felt like good candidates. “Power users” was one — a term we use willy nilly, but without a strong theoretic underpinning. “Network effects” was another, since we were talking about it at a16z all the time, but when it came time to look at the metrics and answer the question — OK so does this product have it!?? — then it got a little mushier. Another was “Product/market fit” or “MVP” and expanding those concepts much further.

The other line of thinking revolves around answering the big question — why? I decided my focus would be on something targeting a very small group of nerdy founders and executives, rather than a wide topic that might be more mainstream. I could write about, say, career advice or how to start a business (in a general sense), but felt like those would be too broad.

In the end, I picked the topic of network effects because it’s a genuinely important topic, I felt like I had something to say, and I also felt like it could fold in a lot of concepts from my prior work on growth. Once I started down the path of picking, I started to talk to people at a16z about it — they recommended I start with writing a book proposal.

Month 1-6: Finding an agent, writing a proposal, and opening up your Christmas presents early
The team at a16z was very helpful, and in the first few weeks I worked with Hanne Winarsky (now at Substack!) and others to start meeting agents, which is how I ultimately met Chris Parris-Lamb from Gernert who also represented Peter Thiel for Zero to One, and Pete Buttigieg for his book. I sent him the following book proposal with a placeholder name, MOONSHOT. The proposal usually kinda reads like a business plan:

  • Overview
  • Chapter summaries
  • The market
  • Author bio
  • Competitive books

We quickly agreed to work together, and that we would approach various publishers to solicit offers. The actual approach was kind of fun, honestly a more efficient version of what we do in venture capital. Chris ran the whole thing, and the process looked like the following:

  • Chris approached publishers and sent along the book proposal
  • They read the proposal (thank you!) and asked for 30 minutes of time
  • We got on the call and they asked me detailed questions, showing they had actually read the proposals — UNLIKE a typical startup/VC process where the founder uses the time to present
  • Later, they submitted an offer (I think 7 did?)
  • Chris then took the top half of the offers, and gave them a second chance to bid again
  • The top 2 bids were close, but I chose to work with Hollis Heimbouch at Harper Business

I chose Hollis because she’s legend in the industry, and worked with Jim Collins, Clay Christensen, Satya Nadella, and others on their most famous books — a16z had also worked with her for Ben’s previous book and it went well. My advance was high mid six figures, which I was told was very good for a first-time author, and would be paid out in parts as the book progressed (one part at signing, the next on the draft, the next at publishing, etc).

The entire process of doing this was maybe 3-4 months? I’ve described the early days of this as “opening up your Christmas gifts early” because you get all the good vibes up front of selling the book, without the work of actually writing anything. But soon I was going to pay the price!

Month 6-12: Collecting and organizing the ideas — lots of fun chats, reconnecting with colleagues, talking to great people
The rest of the first year was pretty fun as well — I realized I needed to do a lot of primary research, so I started reaching out to people I respected, asking them for short interviews. Thank you to Li Jin who tag teamed with me on many of these interviews, where I asked open-ended questions, heard stories, and tried to write as much of it down as possible.

Readers want to hear opinions. The sharper and funnier, the better, and I had a theory that if I could collect all of it, then that in itself could be the bones of a book. Thus, I wrote down pithy, opinionated statements whenever I heard them — anything that might be a good tweet would also be a good title or a good opening paragraph. Opinions like, “launching with Techcrunch is stupid” or “never build a social network, it’s just too hard” — those are gold.

All the interviews went into a spreadsheet tracker like this, which linked to individual notes for each, plus a little summary.

In the end, I ended up with 200+ interviews from people in the industry, and pages and pages of opinions and thoughts. It was absolute chaos. But I could also tell there was something interesting in there. I eventually interviewed some senior folks in the industry — the founders/CEOs of Slack, YouTube, Twitch, Tinder, Dropbox, Zoom, Linkedin, and may others — those all ended up being super fun, and were the showcase stories in the book. Getting time with these folks ended up being some of the most memorable moments while writing the book.

Month 12: How to write the initial the outline, then the mega-outline — finding the formula
If you have hundreds of pages of random notes from interviews, plus pages of research, and a jumble of ideas in your own head — what do you do? You need some kind of organizing principle that makes all these ideas readable. I figured there was probably a formula in some of the best business books out there, and so I re-read Lean Startup, Crossing the Chasm, Innovator’s Dilemma, and many others.

What you find it that the bones of the book often look something like this:

  • Opening story
  • Describe a big problem/dilemma/question
  • Present a framework
  • Go through one part of the framework
    • Start with an anecdote
    • Then describe the theory
  • Go through another part
  • Then another part
  • Then again…
  • Conclusion

This isn’t all business books, but look, it’s pretty ubiquitous. And so I thought I’d start by structuring my initial outline kind of like this, which is how I ended up with the following short version of the outline. The first book outline.

Btw, Ryan Holiday has a great discussion of how he wrote his book, with tons of photos, and I want to link that here. He has a photo of a box representing every topic/idea in this book, each one in a note card, categorized into sections:

I sort of ended up doing the digital version of this, where I created a document that I called my “Mega Outline” — where I took every opinion/point that I wanted to make in the book, and built out the first 2-3 levels of bullets in a much larger version.

Here’s the first page, so you can get a sense:

I’ve linked the entire Mega outline here if you want to peruse — it’s 30 pages where each page needed to probably be 10x’d. That is, 1 page of outline = 10 pages of written prose, which I quickly figured out as I began to write the first few chapters. There’s a funny George RR Martin discussion (he’s the Game of Thrones guy) where he talks about how some writers are Architects, and some are Gardeners. The architects do what Ryan Holiday and I both do — we have some chaos at the beginning, which we try to ruthlessly suppress, and use some organizing principles to put it together. Once there’s a structure, then that’s like a foundation of a building — the architects then write, floor by floor, and build the whole thing. Plus some polish at the end. It turns out that GRRM describes himself as the other archetype, the gardener, where you sort of plant some interesting points here and there, then revisit them as you write. But that’s why his books are amazing and take 10 years to write.

Month 12-24: The very messy middle, the trough of sorrow, the hard slog, followed by trench warfare (yes, it’s 3-5 years to write a book)
This whole middle section after the first year gives me PTSD so I won’t dwell too much on it, and just cover the lessons learned. The mechanics of this phase are pretty simple — you really just need to translate the mega outline bullet by bullet into pages of written prose. But here are all the problems you’ll face:

  • Your normal tools are not good for writing a book. Most writing that you do on a daily basis, like email, might be composed of a few paragraphs. That’s easy. If you need a longer document, then you might have multiple sections that contain multiple paragraphs each, and you’ll use Microsoft Word or GDocs. But what if you have a book with 7-10 parts that contain 5-10 chapters each, that contain 3-4 major sections that themselves contain a large number of paragraphs? And what if halfway through, you realize all the stuff you’re writing in one section should actually belong as a chapter in another section? Also what if you want to do a word count of different chapters or sections? It’s all a pain. In the end, I used Ulysses which at least has the concept of nested folders, and then each chapter would be a folder that would then contain files containing each part. The app then sync’d it all to a bunch of Markdown files in a Dropbox, so that I could work on it from multiple locations
  • You write on a computer, and your computer is very distracting. You need a browser to do research, but your browser is also where you can check what’s happening on social media. You can’t fully turn off the internet, since you need to do research. And sometimes you need to go to YouTube to watch an interview, but right next to the video you’re supposed to be watching is a gadget review for something you might want to to buy. So what do you do?
  • Distraction free devices and treating yourself like a kid. Eventually I started to try and buy a bunch of different tools to keep myself focused. I bought a plexiglass timer safe thing and I’d lock my personal phone away for an hour or two at a time. I bought a separate laptop, and put it in a different location, and turned on all the child-safe filters so that I couldn’t go to Reddit, Twitter, etc. If I needed to look up research, I would often just print out pages and pages of it, so that I would stay analog and not mess around. I bought a series of e-ink Android tablets called the BOOX that could run a Markdown editor, connect to Dropbox, and could pair a nice keyboard.
  • Say goodbye to vacations, weekends, and evening time. To hit the deadlines I had set for myself, I ended up converting a lot of my holidays and weekends into writing time. It’s hard to write for more than, 3-4 hours in a row, so you still can go somewhere nice and sunny — but I found that I needed to wake up, work out, and get writing before noon, in order to make progress. You get your evenings, but it’s tough. And weekends are like that too.

Here’s a funny photo of one of these kSafe timers I’d hide my phone into during my writing times — by the end, I had 5 (!!!) of these in various writing spots, so that if I was feeling in the mood I would throw my phone in:

I have to admit, it was a grind. Not easy at all. If there was a point where I could have gotten stuck and quit, this would have been it.

Month 24-36: Why you’ll feel insecure about the creative process
One of the craziest things about writing a book is that it’s such an incredibly solitary experience, and there’s eventually a point where you’ve written enough that you feel sorta okay about where it’s going, but no one else has seen it yet. And so it might suck. But you’re honestly not sure. I got got to this about 2 years into writing the book. I had written the first ~10 chapters (out of 35), and I had a lot of questions for myself:

  • Is this book any good?
  • Am I saying stuff people already know?
  • Or is this book too nerdy, and going into details that are unnecessary?
  • Are the stories actually interesting, or too obvious? Have people heard them already?

And to be honest, you kind of don’t know until you take a half completed version of the book and ask a few trusted friends to read it. I got a bunch of very very good feedback — thanks in particular to Lenny Rachitsky, Sachin Rekhi, and many folks at a16z for taking the first crack — and it was also the first time my publisher and agent were reading it. I got a bunch of useful conceptual feedback, for example that the first few chapters felt a little slow to get into the action. It felt too theoretical at parts. There were certain specific topics that felt trite. Some sections felt repetitive. And so on. Brutal honesty is what you need here. In the end I also felt like, underneath the scruff, was a book that I would really enjoy reading myself, and that it just needed to be tightened.

I will say, the most painful refactoring happening in this period. As I neared completion of rough versions of all the various chapters, I ended up with a roughly 100,000 word book (which is normal, turns out). Sometimes it takes 3-5 years to fully get to this, and the fact I had a demanding day job and was able to finish in ~3 years — that’s great. But if my worry was that if I had to significantly rewrite portions part way through, it would become a 5 years process, which I’ve learned is not uncommon. This kind of refactoring particularly comes when you have a full length book and then you decide to combine a chapter or two. Or to take a theme that’s appearing in a few spots, and make it into its own section. And then you have to update everything in the book so that it flows properly. It’s easiest to do with a blog post, or a document, or something like that, but with a 35 chapter book — that becomes a heavy lift. But so it goes.

Final months: Just ship it already
By the end of the writing process, I was dead tired. Honestly, I got to a point where I was both simultaneously feeling good about the materials, particularly the first few chapters that I had polished up. But also the process was long and arduous and I was ready to just ship it. The problem with books, however, is that they are really developed in a waterfall process for good reason — once you submit the book, and it’s printed, that’s that!

One fun back and forth happened as I started to work on picking the final cover. I worked with a designer who had done a lot of work on Stripe Press books, which I always loved — however, they are boutique operation which gives them a lot of latitude on what can be done, and the designs often had very small text on the cover (after all, the title will be somewhere on the web page in a digital-first experience, right?), or prescribed weird materials. It was a negotiation to figure out what was actually possible.

I also learned that almost all the US hardcover books are printed at one company (crazy???) and here’s an excerpt about that from a Vox article:

Most book printing happens in the US. Books with heavy color printing, like picture books, are sent to China, but in order to keep the cost of shipping low, most publishers do the rest of their printing domestically. That’s getting more and more difficult to manage.

Until 2018, there were three major printing presses in the US. Then one of them, the 125-year-old company Edwards Brothers Malloy, closed. The remaining big two, Quad and LSC, attempted to merge in 2020, but then the Justice Department filed an antitrust lawsuit. Quad responded by getting out of the book business entirely; LSC filed for bankruptcy and sold off a number of its presses. Smaller printers have continued to operate, but the infrastructure to keep up with the demand for printed books in North America is in shambles.

Crazy right? Couple other interesting things I learned at the end:

  • You only need ~10,000 preorders to be a bestseller — far below what it used to be
  • There are tons of books that become best-sellers because people buy many, many copies of their own books — often via anonymous networks of buyers to obscure what’s happening (I did not do this, btw)
  • The US is not actually the primary market for business books, at least by units — it’s China. There’s usually 3:1 ratio of books sold there versus the US
  • The average book has 250-500 books sold in its lifetime (!!!) — and maybe the median is more like a few thousand. But either way it’s quite low

Anyway, as the final months approached, I traded a bunch of revisions with Hollis and her team at Harper Business. Even though I was very tired at this point, I had the incredible help of Olivia Moore at a16z who did a once over at the end, that really polished things up, as well as my agent Chris, and many others. There are way too many people to thank, so I encourage you to look at the acknowledgements :)

It was only in the final months that I started to think about marketing the book. I also have a ton of notes there :) Will share more later. In the meantime, hopefully y’all found this interesting! It was a good 3+ years of my life and there’s finally enough distance to reflect now.

PS. Get new updates/analysis on tech and startups

I write a high-quality, weekly newsletter covering what's happening in Silicon Valley, focused on startups, marketing, and mobile.

Views expressed in “content” (including posts, podcasts, videos) linked on this website or posted in social media and other platforms (collectively, “content distribution outlets”) are my own and are not the views of AH Capital Management, L.L.C. (“a16z”) or its respective affiliates. AH Capital Management is an investment adviser registered with the Securities and Exchange Commission. Registration as an investment adviser does not imply any special skill or training. The posts are not directed to any investors or potential investors, and do not constitute an offer to sell -- or a solicitation of an offer to buy -- any securities, and may not be used or relied upon in evaluating the merits of any investment.

The content should not be construed as or relied upon in any manner as investment, legal, tax, or other advice. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Any charts provided here are for informational purposes only, and should not be relied upon when making any investment decision. Certain information contained in here has been obtained from third-party sources. While taken from sources believed to be reliable, I have not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation. The content speaks only as of the date indicated.

Under no circumstances should any posts or other information provided on this website -- or on associated content distribution outlets -- be construed as an offer soliciting the purchase or sale of any security or interest in any pooled investment vehicle sponsored, discussed, or mentioned by a16z personnel. Nor should it be construed as an offer to provide investment advisory services; an offer to invest in an a16z-managed pooled investment vehicle will be made separately and only by means of the confidential offering documents of the specific pooled investment vehicles -- which should be read in their entirety, and only to those who, among other requirements, meet certain qualifications under federal securities laws. Such investors, defined as accredited investors and qualified purchasers, are generally deemed capable of evaluating the merits and risks of prospective investments and financial matters. There can be no assurances that a16z’s investment objectives will be achieved or investment strategies will be successful. Any investment in a vehicle managed by a16z involves a high degree of risk including the risk that the entire amount invested is lost. Any investments or portfolio companies mentioned, referred to, or described are not representative of all investments in vehicles managed by a16z and there can be no assurance that the investments will be profitable or that other investments made in the future will have similar characteristics or results. A list of investments made by funds managed by a16z is available at Excluded from this list are investments for which the issuer has not provided permission for a16z to disclose publicly as well as unannounced investments in publicly traded digital assets. Past results of Andreessen Horowitz’s investments, pooled investment vehicles, or investment strategies are not necessarily indicative of future results. Please see for additional important information.