Sunday Reads #125: The Fast and the Ferocious.
Sunday Reads #125: The Fast and the Ferocious.
What we can learn from Amazon and Tiger Global. And what we can't.
Hope you’re keeping safe and sane. It is a troubling time in many parts of the world.
Welcome to the latest edition of Sunday Reads, where we'll look at a topic (or more) in business, strategy, or society, and use them to build our cognitive toolkits for business.
If you’re new here, don’t forget to check out the compilation of my best articles: The best of Jitha.me. I’m sure you’ll find something you like.
And here’s the last edition of my newsletter, in case you missed it: Sunday Reads #124: A new blockchain based Internet is coming.
This week, let’s talk about two Goliaths, that are surprisingly nimble. I’m talking, of course, of Amazon and Tiger Global. What keeps them so fast, despite being gargantuan?
And how can we be faster ourselves, learning from them? Not a simple question: you can’t glibly copy Amazon’s processes, and “Poof! You’re as fast as Amazon”.
Here's the deal - Dive as deep as you want. Read my thoughts first. If you find them intriguing, read the main articles. If you want to learn more, check out the related articles and books. Oh, and do subscribe if you haven’t 😊.
PS. A bunch of readers liked the deep-dive into crypto in the last edition. So next week, I’ll talk about it again. This time, the hidden black swan that could crash the market, and take crypto back a decade.
1. To increase speed, remove dependencies.
Working Backwards is a great book.
Written by two Amazon veterans, it explains the processes and systems that made Amazon the juggernaut it has been for the last 20 years.
One of my key learnings from the book was this: To increase speed, remove dependencies.
Begs a question first: Why is speed important?
Because speed is a competitive advantage. Speed is a moat.
As I said in Speed is a competitive advantage,
Turns out, you can generate a moat out of thin air, by simply being fast. By hustling.
Yes, speed can be a lasting competitive advantage.
In fact, as per Elon Musk, it may be THE lasting competitive advantage.
Says the man who’s started four multi-billion dollar companies:
"The most important sustainable competitive advantage is fostering an organizational culture that supports a higher pace of innovation."
Or if you want something more tweetable:
And Amazon, in particular, SMOKES every other company.
Makes you sit up and pay attention. What are they doing??
Amazon wasn't always so fast.
Oh it was when it started, but then as it grew, it slowed down.
Like every other BigCo: employees scale linearly, dependencies scale exponentially. Suddenly, coordination is a drag on the system.
This became an existential problem for Amazon.
We know they solved this problem - today, Amazon is consistently faster than the other FAANG companies. But how did they do it?
Better coordination is the wrong solution.
One of Bezos' most surprising realizations was that communication is a defect. As he said, "We need to eliminate communication, not encourage it."
Why did he say that?
Because better communication is like a band-aid for a bullet hole (channeling Taylor Swift, another eminent monopolist of our time).
Interdependencies are the reason organizations become slow. You need to remove them, not minimize their impact through better coordination.
This epiphany led to the famous "APIs" memo. Steve Yegge talks about in his rant on Amazon vs. Google:
[Jeff Bezos' memo]:
1. All teams will henceforth expose their data and functionality through service interfaces.
2. Teams must communicate with each other through these interfaces.
3. There will be no other form of interprocess communication allowed: no direct linking, no direct reads of another team's data store, no shared-memory model, no back-doors whatsoever. The only communication allowed is via service interface calls over the network.
4. It doesn't matter what technology they use. HTTP, Corba, Pubsub, custom protocols -- doesn't matter. Bezos doesn't care.
5. All service interfaces, without exception, must be designed from the ground up to be externalizable. That is to say, the team must plan and design to be able to expose the interface to developers in the outside world. No exceptions.
6. Anyone who doesn't do this will be fired.
Of course, this only solves the problem of technical dependencies. What about organizational ones?
The fabled "Two Pizza Teams" was a step in the right direction.
"No team should be larger than can be fed with two pizzas". 🍕🍕
This worked for a while, because it removed the interdependencies between teams. For a while.
But soon, Amazon realized that the model had very narrow application.
Worked only for Product and matrix organizations. Not so much in legal, HR, etc.
Needed a strong multi-disciplinary leader for every two pizza team. But such leaders are rare. Even at Amazon.
And the clincher: sometimes your project needs bigger teams 🤷♂️.
If you're limited to two pizzas, either people or projects will starve.
Given all these constraints, successful two pizza teams were rare.
The search for the Holy Grail continued.
The final (and best) evolution: Single Threaded Leaders and Separable Teams.
Single threaded leaders (STLs) and teams are the best way to boost your organization's speed.
“The best way to fail at inventing something is by making it somebody's part time job” - Dave Limp, SVP - Devices, Amazon.
Single threaded leadership helps with ensuring velocity.
"Single threaded" = the project is the leader's full time job. No other responsibilities.
"Separable" = project teams stand alone. They connect with the rest of the organization only via interfaces. Like APIs (not a coincidence).
Why did this work for Amazon?
It removes the constraint of size of the Two Pizza Team.
You can have very large single threaded teams, of hundreds of people (e.g. for the development of Amazon Alexa).
At the same time, it tackles the root cause, and tackles it well. It minimizes interdependencies between teams.
This is how Amazon AWS and the Kindle succeeded.
But, would an STL model work for you?
The book doesn't talk much about this, but really, this is the most important question of all.
Points to ponder:
Do you have a matrix organization? Is it easy to assemble a multidisciplinary team?
Is your organization set up like "APIs"? Do you have a very clear process to get info from different parts of organization? "I can just call this person for this info" is not good enough. "I email this ID, and I get a reply in 48 hours", is getting close.
Do you have even 10 people working on only one thing full time?
If the answer to any of these questions is No, an STL model might be too hard for your organization to adopt.
Interdependencies will never really go away.
If only because the interfaces will still be people, rather than code.
But the larger message holds. As the authors say, the most successful teams at Amazon were the ones that invested upfront in removing dependencies. Pluck them out one by one. Only after that, start working on the problem.
As Bezos loves to say:
Slow is smooth and smooth is fast.
Further reading:
Sunday Reads #121: How about "Move fast but don't break things"?
2. The ferocious Tiger Global.
I saw this tweet from Packy McCormick the other day.
I had three thoughts on it, going from specific to meta.
#1: Tiger's decisiveness is 🔥
It's unbelievable. Very difficult for other funds to compete.
#2: The structure of capital allocation makes it winner-take-all.
When it comes to building companies, the smaller you are, the more nimble you can be.
Of course, the CEO of IBM will retort, "who says elephants can't dance?".
But really, only startups can move fast and break things.
In capital allocation, it's almost the reverse. The small guy faces structural disadvantages at every step.
I've heard this from friends at small funds. It's difficult to compete with Tiger and Sequoia, who are able to give a term sheet for $20M in 24 hours.
And as I've written before, if you're a small VC or angel, you're nothing but a drunkard searching under streetlights.
The economics and structure of angel investing are heavily stacked against success.
Yes, the hardest part is finding the best companies.
Even if you're almost psychic at picking winners, you can only pick winners among the startups you see (i.e., under your streetlight).
And in a power law world (which the world of startups certainly is), one startup makes all the difference. What if it's one you just missed investing in? That one networking dinner you missed. That one week you were on holiday. The unicorn is just 5m away from your streetlight, but completely in the dark.
And it doesn't end there. Even if you do find tomorrow's billion dollar company, so will other, bigger institutional investors. And they will have no remorse muscling you out.
…This happened to me and OperatorVC. One of our startups hit a strong tailwind of growth, and we were super excited.
Guess what - we were forced to sell at 10x, because the bulge-bracket VC (guess who) investing in their Series B wanted to "clear the cap table" as a precondition to invest.
Yes, 10x is not a bad return. But it's not good enough.
#3: Decisiveness (or lack of it) is where self-proclaimed "rationalists" go wrong. Be a little reckless.
As I tweeted in response to Packy:
Applied Divinity Studies asks a provocative question, "where are the successful rationalists?". The answer is similar:
Rationalists tend to accept the Efficient Markets Hypothesis. They look at an industry, think “what are the odds I know more than people who have done this for a decade?” and assume any seeming inefficiencies are just a Chesterton’s Fence.
That’s not what normal people do at all. Normal people look at an industry, they see a gross inefficiency staring at them in the face, and they think “wow, that’s grossly inefficient!”
And then sometimes, they even set out to solve it.
Nate Soares builds on this, in Recklessness:
[They] would fail because they didn't really expect themselves to succeed. They could make themselves work on their idea, while reciting to themselves some story about being risk-loving, but they couldn't get their head into the idea, to the point where they were spending fourteen hours a day working feverishly while plans and paths and strategies dominated their waking thoughts.
There's a fugue state that successful entrepreneurs report entering, which [this] set of people had rendered themselves unable to enter. Somehow, their realistic understating of their odds destroyed their ability to commit.
In one fashion, this makes some sense: they, knowing that great success is likely a lie, cannot fool their innermost self into believing in their own vision, which precludes them from entering the fugue state.
But in another fashion, is silly. What do the odds have to do with your ability to commit? Why is their epistemic state preventing them from entering the emotional state that would most help them succeed?
As he says, be reckless. Understand the odds. Commit anyway.
3. You wanted flying cars, you got 140 characters...
Well, in the Roaring '20s, you get flying cars too!
A flying car just completed a 35 minute test flight between cities. Check it out - the future is here!
A prototype flying car has completed a test flight between two cities in Slovakia, spending 35 minutes airborne.
The AirCar Prototype 1 is powered by a 160 horsepower BMW engine, and comes equipped with a fixed propeller. It transforms from aircraft to road vehicle in less than three minutes.
After landing in Bratislava on Monday, the aircraft transformed into a car, and was driven into the city center by Klein Vision CEO Stefan Klein and company co-founder Anton Zajac
And if you haven't checked out Github Copilot, definitely do so. Mindboggling stuff! All in a day's work for GPT-3.
4. This made me laugh out loud! 🤣
I saw it two weeks ago and I'm still laughing.
Fibonacci Trump is a true work of art. Pure covfefe.
That's it for this week! Hope you liked the articles. Drop me a line (just hit reply or leave a comment through the button below) and let me know what you think.
I just got my first vaccine dose last week; hope you’ve got yours too. Let’s beat this thing, slowly but surely. Until then, wish you good health, safety, and sanity.
PS. I’m slowing down a little bit for the next few weeks. Will be back to weekly soon!
Until next time,
Jitha