Book review - I'm feeling lucky: the confessions of Google employee number 59
I'm feeling lucky: the confessions of Google employee number 59 by Douglas Edwards (Houghton Mifflin Harcourt, July 2011). ISBN: 978-0547416991.
Summary
"I'm feeling lucky" (IFL) covers the internal discussions of Google's early days (until and including the 2004 IPO) from the perspective of a marketing expert who felt marginalized by the engineering-centric culture. IFL's high points include discussions of strategic decisions about competing search products and explanations about how end-users can relate to Google's customer-facing products. The less interesting parts of the book are the frequent personality clashes with other famous Googlers of the time and vague descriptions of (allegedly) salacious events, e.g., office romances and booze-filled retreats. Mr. Edwards had a minor role at Google, and his outsized ego probably exaggerates his contributions.
Overall rating of the book: 4 out of 5 stars. Worth reading for the strategic decisions. Reviewed by the WSJ in [Hickins 2011] and [Price 2011].
Notes
Page 108. Shared desks, overflowing - "Free is too much". The scrappy startup had to cut costs anywhere it could, which meant the office was overcrowded. The author has a picture on his blog showing desks built out of unfinished doors and sawhorses. The theme of being incredibly cheap with overhead budgets continues throughout the whole book.
Page 132. Feeling unappreciated - Urs Hölzle set the tone of being very harsh in evaluating engineering employees and never providing praise, and the author writes about how this caused problems across the whole company. One passage in particular stands out: "The company stacked its payroll with high achievers unaccustomed to going unacknowledged, and despite the stock options and the free food, they often felt underappreciated." While some of this has changed, there are still large communities of xooglers on Hacker News and Reddit who have said that they quit for these same reasons. It's worth noting that these xooglers turned down the additional money available at Google in exchange for non-financial rewards.
Page 141. Will you finish your PhD now? This section is the most detailed part about Larry and Sergey, who famously dropped out of a Stanford PhD program to launch Google. The author has repeated personality clashes with the founders, who never really seemed to care about marketing very much.
Page 142. Googlers MISC - this mailing list was the location of debates about a variety of office topics, discussions about news, and flame wars. It sounds like the Memegen of yesteryear.
Page 150. Ray Sidney took initiative individually. He was the first employee to really leverage Google's unique culture of individuals being empowered to directly change anything that they felt motivated to change. He unilaterally banned IPs for spammers and attacked the ISPs that hosted them.
Page 161. Google competed against FAST for Yahoo's renewal on inktomi. This is the best part of the book because it shows the strategic decisions which drove Google to success.
Page 166. The Google2 source code was a rewrite of the Stanford code (using Perforce instead of CVS). The index was too slow in the original Stanford code, so it had to be productionized. Urs spearheaded the initiative to rewrite it all. Years later, Google3 was created to replace Makefiles with a custom CI-CD system and replace Perforce with an in-house version control system. This part of the book is the most technical, and the tone changes, as if much of it had been written by others.
Page 188. AdWords sounds like Edwards. He named it. The author named several products, and he maintained the text on the homepage until Marissa Mayer displaced him. Clashes with Marissa became more frequent until Edwards had lost most of his authority. Fortunately for him, post-IPO he had a huge amount of money from his shares and options, so he was able to quit to spend time with his family, which he had not seen regularly for years [Hickins 2011].
Page 228. This is fascinating because the terrorist attacks on September 11, 2001, caused the creation of of Google News, which was a news portal much like the competing products that the founders had so strongly criticized before.
Page 268. Details of ad targeting - the ad industry originally focused on how many ads were displayed (CPM), but Google eventually switched to how many ads were clicked (CPC) and then to how many sales occurred ("conversion rate"). The ideas in this chapter were worth billions of dollars, but it certainly wasn't obvious at the time because competitors had already tried all of the easy and obvious ideas.
Page 272. "Ten things we've found to be true" was one of Edwards's longest lasting contributions to corporate culture at Google, including the famous "don't be evil". Most of these ideas seem to have been lost over time. Even within this book, the author later described how "don't be evil" allowed fully evaluating all of the "evil" options in pursuit of growth. Page 310 - "Don't be evil is not the same as don't consider, test, and evaluate evil."
Page 293. Scientology lawsuit - The Church of Scientology sued Google using the DMCA to force the removal from search results of the church's internal documents and photos from a dissenter's website. This section has a hilarious example of "malicious compliance" - Google did remove the documents from search, but it provided a note about what was absent and where to go get it.
Page 324. The origin of 20% time projects was Gmail content-targeted ads. This is fascinating because everyone thought it was technically impossible to accomplish this goal, and yet one individual engineer did it in his free time, which established the precedent for 20% time projects.