< Back to Blog

Perfect is the enemy of good

Over the Thanksgiving holiday, I read “Chaos Monkeys, Obscene Fortune and Random Failure in Silicon Valley.” The book is a 2015 vintage, but a lot of the lessons and take-aways are relevant today. It’s a great read that I would highly recommend; key excerpts are below.

FB’s data isn’t that valuable. “The miserable conclusion was that Facebook, though assumed to be a rich repository of user data, did not in fact have much commercially useful data at all. Social plugin data, despite its ominous and all-pervasive nature, might fall into that same depressing category.” [pp. 7]

There are two jobs. “To paraphrase the very quotable Silicon Valley venture capitalist Marc Andreessen, in the future there will be two types of jobs: people who tell computers what to do, and people who are told by computers what to do. To take the theory further, computation would no longer fill some hard gap in a human workflow process, such as the calculators used by accountants. Humans would fill the hard gaps in a purely computer workflow process, like Uber drivers.” [pp. 25]

Cofounders need to stick around. “The first sign of trouble was an externally visible one, a symptom that any suitably experienced startup practitioner could have detected: nobody from the early days of the company was still around other than Murthy. Every single other cofounder or early employee had left. As Vonnegut wrote in Bluebeard, never trust the survivor of a massacre until you know what he did to survive. And indeed in this case, Murthy wasn’t a mere survivor, but rather an author of said massacre.” [pp. 43]

Stop claiming you’re in stealth. “Incidentally, the fastest way you can indicate your level of startup naivete to a VC (or to anybody in tech), is either by claiming you’re in “stealth” — that is, with an idea so secretly valuable you can’t disclose it — or by forcing someone to sign a nondisclosure agreement before you even discuss it. You may as well tattoo LOSER on your forehead instead, to save everybody the trouble.”

H1B visas are often used as handcuffs. For this last piece of retention skulduggery Adchemy would exploit one critical point: Argyris wasn’t a US citizen. Skilled immigrant tech workers in the United States have effectively one method of entry: the famous H-1B visa.” [pp. 69]

You’ve got to be viral. “Marketing is like sex: only losers pay for it.” [pp.74]

YC insight. “This is how Y Combinator works. For three months, the selected startup founders meet weekly for a dinner with some eminent startup personage. The dinners are not relaxed social affairs: they’re competitive demos in which founders try to one-up each other with increasingly developed products, upward-sloping user graphs, or funding news, within a context of technocamaraderie and shared suffering. The real YC selling points were the following: access to the YC partners, access to the network of YC founders, a bankable patina of prestige, and Demo Day.” [pp. 78]

Google’s moneymakers. “Think about this in the context of more traditional industries for a moment. Chain restaurants like McDonalds have a best-performing outlet in a particularly busy high-rent district. Automakers have a particularly popular, bestselling model like the Ford Fusion or the Chevy Impala that makes their quarter. Google has the words “insurance” and “loans.” Those are its flagship moneymakers. It is a tech empire built on snippets of words and phrases flitting through people’s minds. You as a mere consumer don’t see it, but how Google ranks keywords and runs the auction determines the fate of billion-dollar companies and industries. As the gatekeeper on buying interest, Google is the bouncer at the door of almost any Internet-enabled business today, and if you as the business owner don’t pay your bouncer well, he’ll shut you down, as Google has done more than once.” [pp. 82]

No co-CEO’s or equal partners. “Long story short, AdGrok was a complete basket case when it came to founder dynamics, and it was mostly because we were all equal in equity, and therefore in power as well. What’s worse, since there were three of us, we’d commonly form two-person factions against one dissident. Personally, I’d go through periods where I was tight with Argyris (of course, against MRM), and then back to MRM against Argyris. It was madness. Here’s how it needs to be: either you’ve achieved a certain Vulcan-quality mind-meld with your founders, your brains welded together in the crucible of some formative life experience like the military or hard-won work experience, or there’s one guy running the show (with at least 51 percent of the equity). End of story. The CEO has the last word, despite all the fervid debate that happens in an early-stage startup. Everyone needs to accept that, or absent him- or herself from the proceedings. As [Paul Graham] told us all: Have a leader!” [pp. 90]

Post PR on a Tuesday. According to the leading PR mythologies, day-of-the-week posting choice was critical. The media boom would reverberate depending on its magnitude and resonance inside whatever industry echo chamber it was launched. And so you wanted a few full-on workdays after launch to let that echo play out. Monday was too soon, as everyone was still jet-lagged, hungover, or otherwise groggy from the weekend, and too busy catching up on email and meetings. By Thursday, people were already thinking about the weekend, and likely ducking out early for their first happy hour of the week. Friday was for burying news, not announcing it. It’s when people were fired and bad earnings reports came out. We would post on Tuesday, which left the most time for a PR blowup to echo across the Internet, and across all levels of Internet connedness, from the assimilated Internet cyborg to the grandmother in Kansas. [pp. 99]

Negative engagement is useful. “The post itself quickly accumulated dozens, and eventually hundreds, of comments, positive and negative, and both equally useful. By the end of it, we had thousands of sign-ups. This post [about Goldman] would be the first in a series of hyperviral blog posts that would put AdGrok on the startup map (if not quite the customer one). Every three to four weeks, another gaseous emanation from the latrine of human thought (aka me) would appear and rocket us to the top of Hacker News (the tech geek’s Cosmo), and make another stir in the evanescent tech buzz-o-sphere.” [pp. 102]

Caps aren’t what they used to be. “In the summer of 2010, for the prechosen YC elite, a good cap was in the $6 million range. A stellar cap was around $8 million, and only a really buzzy company like Hipmunk, a travel startup that was a Reddit founder’s second act, got close to that. The middle of the YC pack (where we were) was in the $3 million to $4 million range.” [pp. 116]

Only the CEO goes to pitches. “Now that the boys [the cofounders] realized what we were dealing with, they wanted to ride along to the pitch and see what the great Sequoia was about. This was a bad idea. In general, either the CEO or the smooth-tongued founder designated for the purpose should be involved in the fund-raising, and that person alone. Whenever you face some stressful, time-consuming, and risky challenge, firewall the rest of the company away from the mess.” [pp. 123]

Don’t talk to associates. “One quick way to cut through the shit: ask your pretender-to-influence, “Do you have decision-making power?” If he or she even remotely hesitates or hedges, you’re speaking to a lackey (whether he or she acts like one or not). His or her only utility is to get you to the person who does have that power; everything else is so much pig swill. So route around such people if a real investment check is your goal. Arguably (and this is the canonical YC advice) don’t even accept a meeting from someone who can’t answer yes honestly to the above question. You’re wasting your time.” [pp. 129–130]

Lawsuits are expensive. “All together we were looking at around $1.5 million for the full lawsuit, including judgment and final wrap-up, stretching out over the next eighteen months.” [pp. 141]

YC will have your back. “Y Combinator was the sort of unforgiving power player that remembered the names of investors who had crossed portfolio companies in the past, or who had disseminated unflattering portraits of YC, and blacklisted them from any YC dealings, or from the minds of YC founders. This was done with little thought of the size of their funds or their influence in the Valley, leaving more than one self-important VC sputtering at a dressing-down, or left out of a round because the company’s founders had gotten a warning from PG [Paul Graham]. The harsh reality is this: to have influence in the world, you need to be willing and able to reward your friends and punish your enemies.” [pp. 157–158]

CoFounders keep you going. “In fact, as in sports or war, the overpowering desire to not let down your fellows is often all that keeps you going. But if you have no cofounders, or if that band-of-brothers relationship isn’t there, then the only thing keeping the whole construct going is sheer, stubborn bloody-mindedness. You get up every morning, get kicked in the face, and come back for more the next day.” [pp. 165]

Don’t talk to acquirers too soon. “PG was my first email. As expected, he was bearish on the whole thing, calling it a distraction, and advising us to ignore it and to get used to saying no. He was right, in the sense that any decent startup will get a dozen acquisition offers as it rises to prominence… He also advised me privately to shield the boys from any acquisition chatter.” [pp. 187]

The way acquihires really go. Companies with acquisition wherewithal and the nerve to use it bid for what they wanted in deals. You came in with your team and your product; they gave it the once-over, and said, “We want person A and B, but not C, and we don’t care about the tech.” They then offered you a lump sum for what they wanted, and you were left to double-deal, buy out, or otherwise fuck over whoever in order to get the deal done. [pp. 231]

The other way acquihires can go. As I was soon to learn, this is a common tactic. You pay a pittance for the company, engineering it such that investors get little, and then pack the real value into the hiring offers for the employees. TechCrunch would carry the news that the company sold for X million dollars, but technically it would sell for 10 percent of X, while the rest went into fat signing bonuses and heavily laden vesting schedules for the founders. [pp. 245]

Google & FB competed for the best talent. Companies are like countries: the populations really vote only with their feet, either coming or going. Google instituted a policy whereby any desirable Googler who got a Facebook offer would have it beaten instantly by a heaping Google counteroffer. This, of course, caused a rush of Googlers to interview at Facebook, only to use the resulting offer as a bargaining chip to improve their Google pay. [pp. 287]

Perfect is the enemy of good. In general, be it at startups or aggressive companies like Facebook, there should be a cultural bias for launching. The perfect is very often the enemy of the good, and as the Facebook poster screamed from every wall: DONE IS BETTER THAN PERFECT. Very few companies have died due to launching early… [pp. 294]

The best orgs are metrics driven. “A metrics dashboard that the entire team fixates on was standard Facebook practice. Those scorecards should obsess the product manager, should be the last thing he thinks of at night and the first thing he thinks of in the morning, and should be known to him, from memory, down to the decimal. Choose well what goes in that line graph, because whatever it is, a good product manager will slave away until it’s going up and to the right, whatever that figure represents. You make what you measure, so measure carefully.” [pp. 296]

FB’s data isn’t as valuable as you think. “No user data we had, if fed freely into the topics that Facebook’s savviest marketers used to target their ads, improved any performance metric we had access to. That meant that advertisers trying to find someone who, say, wanted to buy a car, benefited not at all from all the car chatter taking place on Facebook.” [pp. 296]

NZ is a testing ground. “As a geographic tangent: New Zealand was commonly used as a test bed for new user-facing products. It was perfect due to its English-language usage, its relative isolation in terms of the social graph (i.e., most friend links were internal to the country), and, frankly, its lack of newsworthiness, so any gossip or reporting of new Facebook features ran a low risk of leaking back to the real target markets of the United States and Europe.” [pp. 318]

The right way to be a product manager. “The principal reason for you to be technical is not to help technically design the system under development; if you’re doing that, then you’re PMing wrong. No, you’re technical so you can tell when engineers are bullshitting you, which will be often. At times it’s accidental (as it was with Rong), due to either miscommunication, bad memory, or wishful thinking (engineers are as inclined to it as anybody). Sometimes it’s more stealthy, their passive-aggressive way to disagree with the product direction (“That’ll eat up all our servers”), or laziness (“It’s impossible to build that”). The PM is there to give a sniff test to any such product-killing assertion.” [pp. 321]

Ads are a necessary evil. “Advertising is the only reliable business model that’s worked for all but a handful of publishers, and those only among the elite, content-producing ranks of The Economist and the Wall Street Journal, who manage to charge their users. If you want to interact with the world via the Internet, then deal with ads.” [pp. 325]

FB isnt as smart as you think. “The story usually ran as follows: the journalist, or one of her “sources,” had seen ads for the San Francisco 49ers after her husband’s cousin’s college roommate posted a photo of himself in a 49ers jersey. Were we using uploaded photos in our ads targeting!!?? This was like being accused of fathering Scarlett Johansson’s love child. I wish I could even reasonably be suspected of pulling that off. Almost nothing of what you share on Facebook… is worth anything in commercial terms… Facebook doesn’t sell your data; it buys it. It does this by providing services to advertisers that incentivize them to let Facebook ingest the data you’ve generated outside Facebook.” [pp. 328]

You can go extinct. “When we moved into Menlo Park, there were Sun logos on lots of conference room doors and public spaces. Rather than remove them all, Zuck ordered that a few of them remain. Like corporate memento mori, they were to remind employees that Facebook could also go the way of extinction, and be reduced one day to logos and swag.” [pp. 337]

Be your own creative destruction. “The penultimate page captured the spirit best. In white sans serif font, against a stark black background, it read: If we don’t create the thing that kills Facebook, someone else will. “Embracing change” isn’t enough. It has to be so hardwired into who we are that even talking about it seems redundant. The Internet is not a friendly place. Things that don’t stay relevant don’t even get the luxury of leaving ruins. They disappear.” [pp. 343]

Startup guy will get bored and wont fully vest. “As early as a few months into Facebook, once the novelty had worn off, I could feel myself growing bored and frustrated with the speedometer stuck in the middle double digits. Product development in Ads was sluggish and curiously hesitant. My one major oversight was failing to realize that the impatient startup entrepreneur, the guy addicted to the smell of burnt clutch and the engine’s throaty roar, was never going to stick around for four years and vest in peace (something the corp dev people who put together my offer surely realized, those bastards). You were really going to spend at best two years there, so halve the number on your offer or term sheet, and then halve it again if it’s in the form of stock rather than options (thanks to ass-backward IRS treatment of Valley compensation).” [pp. 347]

Engineer salaries in 2014. “If Facebook engineers coming in right before the IPO had a few years of experience, but were not a “star” with unique skills, they could expect a salary in the $125K to $150K range, with about $500K in equity.” [pp. 358]

FB is running out of users. “Without China and Russia, and taking a 25 percent haircut of people who’ll never join or stay (as is the case in the United States), that leaves around 1.8 billion potential Facebook users globally. That’s it. In the first quarter of 2015, Facebook announced it had 1.44 billion users. As such Facebook can’t wait for the developing world to get to First World standards of connectivity, so it must create it for them, using ad revenues in the developed world to subsidize this new air force’s deployment. In time, monetization will follow usage, as it always does. Money follows eyeballs, even if slowly.” [pp. 376]

Mobile Newsfeed saved FB. “That’s it: ads in News Feed, while the user was on his or her mobile device — that’s what saved Facebook. Like many successful products, News Feed Ads rode to success atop a tsunami-esque wave nobody had predicted, or at least hadn’t predicted to arrive right then and so quickly. In this case, that wave was mobile usage, which in the span of a few months in 2013 suddenly constituted the majority of Facebook usage.” [pp. 483]

Mobile world means less data sharing. “In mobile, Web browsers generally don’t accept third-party cookies, which means that someone other than the New York Times can’t read or write data about you when you’re on nytimes.com on your mobile browser. Again, compare that with desktop, where you have Chrome or Safari (or one of a handful of browsers), with which you browse thousands of sites, and all the data sloshes together across your browser’s cookie pool, getting siphoned off and resold in a bajillion ways. In mobile apps, that data mosh pit doesn’t exist, as all the data is siloed within the app that generated it. If you’ve gotten to level 47 in Candy Crush Saga, searched for a house on the Redfin real estate app, or bought something on Amazon’s mobile commerce app, that data lives and dies inside those apps, and never leaves.” [pp. 484]

Targeting is less valuable with mobile. “The net of all this detailed data discussion is this: while my statements about the questionable value of Facebook data held true on desktop, which already had a mature and mostly respectable data marketplace by the time Facebook showed up peddling Likes, that was emphatically not the case on mobile. On mobile, targeting data was sparse, and bad when it did exist, such that even basic targeting like age and gender was a godsend to data-starved mobile marketers who’d been mostly shooting in the dark.” [pp. 486]

Mobile is the enemy of most publishers. “All this meant that while on desktop high-quality publishers with engaging formats competed with Facebook, on mobile the pickings were very slim indeed. Nonintrusive but stylish ads, which paired well with organic content from your friends, on an app experience like Facebook, which featured supremely focused attention and a crazy-high engagement rate (clickthrough rates on Facebook’s News Feed reached easily into the single-digit percentages), were very competitive with the mobile alternatives.” [pp. 487]

FB’s IPO was brilliant for FB. “Also recall, Facebook’s IPO, unlike Twitter’s, came out at a high price of $38, and then languished for a year around $30, occasionally going so low as $18. The IPO was great for employees and insiders worried about dilution, but not for people wanting to cash out (like me), and walk away from the Silicon Valley casino.” [pp. 496]

Chaos Monkeys is well worth the full read.

Visit us at blossomstreetventures.com and email us directly with Series A or B opportunities at sammy@blossomstreetventures.com. We invest $1mm to $1.5mm in growth rounds, inside rounds, cap table restructurings, note clean outs, and other ‘special situations’ all over the US & Canada.