Burning the Page

by Jason Merkoski

  So the first time I got a Kindle, it wasn’t called a Kindle but a “Fiona.”

  Though primitive by today’s standards, my original Kindle—one of the first Fionas made for select Amazon employees—still works like a charm. True, my Fiona is turning the yellow-gray color of smokers’ teeth, the same way that once white yesteryear computers start to turn an upsetting beige. But it still works, even though it’s been manhandled and chucked many times into my backpack, tossed into many suitcases for trans-Atlantic flights, and left on my truck’s dashboard in the sun for months. And once while walking through Cupertino, California—a city where everyone drives—I got hit by a car while crossing the street, because nobody expects pedestrians in the heart of Silicon Valley. I fell and sprained my arm. But even though my Fiona clattered to the street and got run over by one of the car’s wheels, it still works as great as always.

  Needless to say, I love my Kindle.

  My original Kindle job had me creating and managing the ebook conversion process—the messy method by which print books are turned into digital ones.

  When thinking about how ebooks are created, it’s best to envision a sausage factory. Meat comes in one end, machinery packages it, and sausage comes out the other end. At the ebook factory, you start in the front with books from publishers. They’re chopped up, reassembled and packaged, and finally made available for sale in digital form.

  Most ebooks are created using a digital copy of the physical book, usually in PDF format. PDF files have a fixed layout, which means they’re formatted in the way they’re supposed to appear on a printed page. However, ebooks need to be reflowable, which means that if you change the font size on the ebook, the words and sentences and paragraphs should be reformatted so that the words wrap around properly in the paragraph. You can’t do this well with PDFs.

  To make a PDF into a reflowable ebook, publishers usually use a conversion house. Such companies, in turn, use a combination of software and workers overseas. Many of the conversion houses use people in India or China, or sometimes more exotic places like Sierra Leone or Madagascar or the Philippines. They usually work in a large warehouse or an old factory, with cubicles running from one end of the factory to the other on multiple floors.

  Elbow to elbow, the workers stare at words on the screen all day, reading ebooks. They remove page numbers, reformat the ebooks to make them reflowable, and skim through them afterward to make sure no paragraphs or illustrations from the originals were lost during the process.

  But not all books are in PDF format; some only exist in print. More brutal methods are often needed to digitize such books. As part of my job, I got to watch as workers destroyed print books to turn them into ebooks. Pages had to be removed from books so they could be scanned and digitized. As a book lover, I was horrified. To remove the pages of the book, workers would hack the spines off with knives like they were whacking their way through the jungle with machetes. Once their content was scanned, those pages would be tossed into a Dumpster at the end of every shift.

  It was destructive, and the books could never be recovered afterward. The ebook revolution was bloodless, in the sense that there were no human casualties. But if books could bleed, you’d find their graveyards overseas. You’d find burial pits, unmarked graves, and hundreds of thousands of casualties.

  But all this was needed to launch the Kindle; we couldn’t just launch a hardware product without any ebooks to read. Without ebooks, the Fiona device would have been just an expensive paperweight.

  You see, we needed both the ebooks and the hardware for the Kindle flywheel.

  Many people in dot-com and tech companies think in terms of “flywheels,” but most nontechnical people don’t know what that means. It probably sounds like lots of flies strung up to a mill wheel, slowly turning it to crush wheat into flour.

  In tech terms, a flywheel is something that builds up energy as it spins. The goal is to get it spinning faster and faster, however you can. The faster it spins, the more energy you have (or in business terms, the more money you have). The Kindle flywheel, for example, might start with launching an e-reader into the marketplace with a small number of ebooks. People buy the device, and then they use it to buy ebooks. The profit from both can be used to build an improved e-reader, which can be sold more cheaply, which then means more people will buy it and consequently buy more ebooks, the profits of which can then go back into building even better, even cheaper Kindles. With every push the flywheel gets, the faster it spins and the more powerful it becomes.

  The Kindle flywheel started spinning fast as the Kindle business grew. And in true Amazon tradition, the business was run with metrics, with meetings called “deep dives” where the team would dive into spreadsheets. Amazon is a highly numerate culture. The numerically literate seemed to do well there, because they could mentally pivot rows and columns of spreadsheets and crunch numbers on the fly.

  During a deep dive, you let go of preconceived notions and think logically. You look at data—instead of doing a technical hand-wave, you speak to the specifics. In Amazon’s deep-dive culture, facts are preferred to opinions. Deep dives are like science experiments, and you approach them with a hypothesis you want to prove. If your hypothesis is disproven, then you come up with a new hypothesis, run tests to gather data, and analyze data to prove or disprove the new hypothesis.

  Most of the engineers at Amazon dreaded these deep dives because they had to put on something formal, like a button-up shirt and a pair of jeans with a belt. Amazon isn’t a formal place: a J. Crew shirt and Dockers are as formal as it gets. But still, for engineers, even wearing these is an affront against nature, a blasphemous abomination out of a Dungeons & Dragons game or an accursed H. P. Lovecraft story.

  In one of my first meetings with Jeff Bezos, we were doing a deep dive on ebook content and what it looked like on the Kindle. We sat and used our Kindles as customers might. In some ways this was like the first digital book club; we were mostly silent, just reading on our Kindles. Sometimes we would annotate content or buy a new book—anything to test all the features.

  At one point, Jeff’s Kindle must have crashed, because it became unresponsive. The room had been silent for a while because we were all absorbed in our books. Then out of nowhere, Jeff exclaimed: “I’m hung! I’m hung!” I looked up with a surprised grin on my face, but Jeff was unaware of his double entendre.

  All the others in the room were actively trying to stifle their laughs. There was a little bit of hero worship at Amazon. Now, I admire anyone who runs a bookstore, so I can’t help but admire Jeff Bezos. Not only does he run the world’s biggest bookstore, but heck, he has his own rocketship company too. But some of my colleagues took admiration to a whole new level.

  I don’t think anyone at Amazon deliberately shaved their heads bald to look like him, but people would be in a Jeff meeting and come out afterward and rave about Jeff’s stories, how he laughed, or a savage insight he had. People would find out about the books he was reading and read them too. (During the Kindle years, The Black Swan: The Impact of the Highly Improbable was popular among the Jeffnosanti, although a book he read on the history of tungsten was slightly less popular.) People routinely lionized Jeff for how much money he had and his high IQ. So they certainly did not want to look like they were laughing at him or criticizing his ideas.

  Let’s face it: we all contributed to Kindle, but Jeff was the visionary, and digital books will be his legacy. True, there were other digital book pioneers. Heck, I was one of them. I made the first modern ebook in 1999, and I invented my fair share of Kindle features. And I wasn’t alone; we all invented Kindle in our own ways. None of us who toiled in the Kindle workshops were flunkies. We were all colorful characters, innovators, and pioneers.

  But only Jeff had the vision and the millions of dollars in seed capital to start Kindle. And trust me, it took a lot of capital, considering the salaries and stock grants for the employees the first few years, as well as all the R&D and acqu
isitions and startups he had to fund. Jeff not only saw the dream; he also made sure the dream happened, at great financial risk.

  So as difficult as our challenges could be, life at Amazon felt like we were creating something revolutionary, and we had the financial means to do it. We were like techie versions of the early workers who toiled in Gutenberg’s workshop.

  Life in the Kindle offices in those early days was like working in an alternate, over-caffeinated, sugar-high universe. And I loved it. The offices were loud, with the sounds of BlackBerries and pagers going off. The building shook every ten minutes as a streetcar rumbled past, and the hum of a microwave melting someone’s leftover Indian dinner filled the air at lunchtime. Inevitably the cries of an engineer shouting at the top of his lungs would emerge from a conference room, along with the pounding of his fists against the whiteboard walls.

  In the kitchen you would find occasional stacks of Top Pot donuts, local Seattle fritters that tasted like they’d been deep-fried in nothing but pure sugar, cocaine, and aspirin. You’d also find the remains of catered breakfasts or lunches that senior management would put out in a kitchen for anyone else to have when they were done eating, like lords of the manor throwing their serfs an occasional bone to nibble on.

  Like most technology companies, Kindle had lots of beer, usually on Friday afternoons. People would often bring in six-packs, open them near someone’s desk, and stand around and talk at the end of a long day. Some crazy conversation would emerge, full of crazy, hypothetical what-ifs like: “If you could suspend a killer whale from a rope and suspend a tiger from a rope and let them attack each other, who would win?”

  “Decoration” was hardly the word for what you’d see in the warren of cubicles we inhabited. If you walked among them, you would find Amazon-issued Magic 8 Balls, humming computers, Kindles connected to power cables, teetering printouts of architecture diagrams or spreadsheets, posters from Battlestar Galactica on how to spot Cylons, tipsy engineers still arguing about whether the killer whale would get the tiger, discarded Kindle boxes being used to prop up Foosball tables, and an arcade-style, fully functioning Donkey Kong game I could never beat.

  Amazon was, in short, a bit of a sloppy Seattle dot-com—but one with billion-dollar revenues and razor-thin profit margins. Those thin margins meant that we had to stay focused on launching Kindle, without distraction.

  » » »

  Secrecy was important in the early days of Kindle. We weren’t allowed to take our Kindles home or show them to our families or get caught using them in public, out of fear that someone outside of Amazon would see the Kindle and leak information to blogs or newspapers.

  But with this secrecy came a great feeling of pride and privilege. I felt like one of the first people to use an iPod, years before anyone else even knew it existed. The Kindle was a secret I couldn’t share with anyone, not even my family!

  Until the Kindle launched, the only other place on the planet that knew about it was Lab126 in Cupertino, California, where the Kindle hardware was designed.

  In the very early days of Kindle, when its eInk screen was just a gleam in Jeff Bezos’s eye, Amazon was smart enough to realize it had never done manufacturing before. It was great at website sales, but it had no expertise in making hardware. Jeff decided it would be best to spin up a new organization solely responsible for this.

  The name Lab126 came from a technical kind of pun. Amazon already had its “A to Z” development center in Palo Alto, which developed technologies like the A9 search engine that Amazon uses. Jeff wanted Lab126 to be a research facility—hence the “Lab” part of its name.

  As for the “126” part, well, you have to realize that there was never a Lab125 or a Lab124, just like there was only ever a Preparation H, never a G or an F. The “126” part stems from the fact that “A” is the first letter of the alphabet and “Z” is the 26th, a techno-geeky homage to the “A to Z” development center. Jeff liked his geeky in-jokes—you could have heard his laugh a mile away when they came up with that name.

  To attract and retain the best hardware engineers, Lab126 would be located not in Seattle but in Silicon Valley. The Lab126 offices were originally in a mini-mall across the street from a music studio and a slightly sleazy jewelry store. But due to the number of new hires, the lab quickly grew out of its old space and moved to Cupertino, right in the heart of the Valley. Moving to Cupertino put them in the big leagues—they were now in the same city as Apple.

  After a year running Kindle’s ebook software team, I was asked to take the lead in launching Kindle as its program manager. That meant I had to know everything about the Kindle hardware, so I started flying down to Lab126 on a regular basis. I went to bridge the gaping cultural chasm between Amazon and Lab126. Everyone in Cupertino understood hardware, and everyone in Seattle understood the web, but neither understood the other. Amazon understood web services; Cupertino understood consumer electronics.

  And the combination of the two: ebooks? It was new territory for everyone. Almost no one in the company had exactly the right set of qualifications to help Lab126 and Amazon speak to and understand each other, with one exception: me. I used to work at Motorola making cell phones and internet routers, so I could speak the language of hardware people, but I had also built websites for companies like Home Depot and Walmart, so I could speak the language used at Amazon.

  On visiting Lab126 for the first time, what you’d notice would be the stark contrast in floor layout between Amazon’s offices and Lab126’s. Amazon has a messy organic layout. All the floors are open, with people at desks sitting side by side in a vast room without walls, like a Southeast Asian call center or some fly-by-night dot-com’s tech support division. The Lab126 offices resemble a printed circuit board, in keeping perhaps with the mentality of a hardware engineering company. All the cubicles and hallways are aligned at right angles, with efficient pathways in between. Being at Lab126 was like being on the circuit board inside a Kindle.

  » » »

  Much as I wanted to read on my Kindle while flying on a plane every week between Lab126 and Amazon, I couldn’t because Kindle was still such a secret. I couldn’t even bring the Kindle through airport security, in case the security personnel needed to examine it. Plus, I feared that journalists or competitors might see my Kindle in the few seconds that it would be in plain view.

  I spent two years traveling back and forth between Lab126 and Amazon. When you’re a kid, the years have a way of passing quickly. All you can seem to remember when you look back are summer nights, fireflies, and snowball fights. The same was true of me with Kindle. When I look back at the years leading up to the Kindle launch, it’s like I was a kid, moving happily from one day to the next, one challenge to the next.

  One of the challenges required Jeff’s personal attention and had to do with the Kindle ebook format. Nobody else on the Kindle team believed it was important enough to merit his attention, but I did, so I set up a meeting with Jeff to discuss it. (I suspect that the days when someone can set up a meeting willy-nilly with Jeff are over, now that Kindle has grown so large.)

  Now, just because you had set up a meeting with Jeff didn’t mean it would actually take place. To get to Jeff’s office, you had to get past his executive assistants. They have offices of their own, and in a Kafkaesque way, you’d have to talk your way past the first executive assistant to see the next one and then talk your way past her to make your way to the third assistant, and so on. Eventually you got to Jeff’s office, where you’d probably find him gone and realize that they’d neglected to say he was out for the day.

  The day of the meeting, I made it to Jeff’s office a little early, before he had arrived from another meeting elsewhere. I looked through his windows and tried to understand the way he saw things. He had a telescope in his office and pictures of his kids on the wall. It was a small office, actually, dominated by a giant work desk with tidy stacks of papers.

  I imagined him looking out through his telescope at h
is far-flung workers, spread out as they were through Seattle in different office buildings, and I imagined him perhaps aiming his telescope at his fulfillment centers in Kentucky or Nevada, yearning to see the incessant shipments of everything from books to Beanie Babies, DVDs to diapers.

  Protected by his executive assistants and sequestered in a tower in Amazon’s headquarters, Jeff’s office was a little like a walled garden. It’s an appropriate metaphor, because what Jeff and I discussed that day, and on days and weeks to follow, had to do with Kindle’s own walled garden.

  When you’re reading about companies like Amazon and Apple, you often come across the “walled garden” metaphor. I want to explain it to you with a visual metaphor, because I’m a visual guy.

  Imagine the wall of a medieval fortress. There might even be a moat around it. It’s a tall wall, made of stone—a wall to keep the enemy out. There’s one way into and out of the fortress, and that’s over a drawbridge that comes clanking down to let you across the moat, through a hole in the wall, and into the city inside. You can think of the city as being everything good that the wall is supposed to be protecting—all the people and gardens inside. This wall protects you from the dragons outside, from the Vandals and Huns and would-be conquerors.

  In tech terms, the walled garden is the arrangement of software and hardware and file format that makes it almost impossible to get to what’s inside unless you go over the drawbridge, the officially sanctioned way in.

  Look at the iPod. It relies on a proprietary format, a proprietary way of getting content into and out of the device. And yet it’s successful because the walled garden is tended so carefully.

  Amazon has a similar walled garden for the Kindle. The only way you can buy a book and read it on the Kindle, according to Amazon’s walled garden approach, is to buy the book from the Kindle store. Are there other ways of reading a book on a Kindle? Yes, but they’re equivalent to the Vandals and Huns laying siege to the city by running ladders up its ramparts and then climbing those ladders with axes and grappling irons. In modern tech terms, this kind of attack is piracy. Or if not outright piracy, it’s that gray area related to digital rights management (called DRM)—the restrictions used to keep people from copying or sharing ebooks for free.