Uncompromising Celerity
Building ambitious things fast
“There is no secrecy comparable to celerity.”
-Francis Bacon
One of my favorite pages on the internet is owned by Patrick Collinson, the CEO of Stripe. He hosts it on his personal website here. The title is simply 'Fast'. The page catalogs a bunch of projects that were highly ambitious and accomplished very quickly. It is said that when you have a project, you can choose two of the following: have it cheap, have it high quality, or have it fast. In rare instances you can get all three. For example:
The Empire State Building. Construction was started and finished in 410 days. Source: Empire State Building.
Once completed, the Empire State Building was the tallest building in the world for nearly 40 years. It required new construction methods and was built faster than any other tall building of the time. In fact, it still holds a record on a floor-by-floor basis. Burj Khalifa, the current tallest building in the world took 2,095 days to build at a rate of 12.8 days per floor (for 163 floors). Compared to the 4 days per floor (for 102 floors) of the Empire State Building. Burj Khalifa took about 3x longer. Interestingly, there were roughly 3,400 laborers working on the Empire State Building and about 12,000 laborers for Burj Khalifa. So nearly a 10x difference in output.
Similarly, look at how long it took to create the Visa card:
BankAmericard. Dee Hock was given 90 days to launch the BankAmericard card (which became the Visa card), starting from scratch. He did. In that period, he signed up more than 100,000 customers. Source: Electronic Value Exchange.
One hundred thousand customers in ninety days from scratch is an incredible feat and one that still holds up today as an elite achievement. Even ChatGPT took longer from scratch to reach 100,000 customers, and it reached over 100 million customers in 2 months after launch, following several years of development.
Well you might say, these two projects are really just projects, not a fundamental technology that's used in other inventions. Fundamental foundational technologies surely must take longer to develop. Now let's look at Unix, which most modern software applications are built upon:
Unix. Ken Thompson wrote the first version in three weeks. Source: UNIX: A History and a Memoir
Very fast indeed. Apparently world changing technologies can be created very quickly. Granted they still might be continuously iterated upon.
To contrast all these fast projects, Collinson relays a slow infrastructure project that would not surprise anyone, which tends to be par for the course these days:
San Francisco proposed a new bus lane on Van Ness in 2001. Its opening was recently delayed to 2021, yielding a project duration of around 7,300 days. “The project has been delayed due to an increase of wet weather since the project started,” said Paul Rose, a San Francisco Municipal Transportation Agency spokesperson. The project will cost $310 million, i.e. $100,000 per meter. The Alaska Highway, mentioned above, constructed across remote tundra, cost $793 per meter in 2019 dollars.
Different
When I was younger, one of my mentors taught me to go after big things because you’ll spend about the same amount of time working on it but you’ll accomplish more. In a similar manner, when studying at school, I would always focus on being able to do the hardest, most complex problems for a topic because if I could do those then everything else would be easy.
There’s a mindset shift when you start tackling big, ambitious things. It opens up new ways of thinking and every obstacle overcome builds momentum that pulls you forward. Every great video game player and designer understands this feeling. When you look at the projects above, there were innumerable projects similar in scale that took much, much longer to complete. You need to ask what changed in how these people thought and acted to make them accomplish something so big at a much faster pace.
How can one have outsized performance? It's not a single thing but a confluence of many factors. A mindset with the right types of actions and systems. Anyone can do one of the following well, but doing all of the below consistently provides outsized results.
Ownership - a true feeling and assignment of ownership up and down the chain always ends in a better end product. A great book on this is It's Your Ship by Captain Michael Abrashoff.
Building for the next two steps - doing non-critical path items that make future critical-path steps go faster are worth doing and can make future progress snowball.
New innovative methods - lots of people innovate and still don't make improvements. Looking at the Burj Khalifa vs Empire State Building example, one would think 70 years of technological progress would create improved building methods but apparently not. However, you can't get anomalous results if you do things the same way everyone else does.
Ambitious goals - big goals, especially those that haven't been accomplished before can create gravity and momentum as teams march towards their accomplishment. Taking down Goliath has a way of rallying people to finding alternative, better solutions.
Technically strong/masters of craft - great knowledge from lots of experience allows you to do things that others can't. At the same time, knowing what can't be done and what ends in failure enables you to avoid solutions that won't work. There's a lot that compounds from just avoiding bad decisions.
Provision for randomness - in large, ambitious projects things are going to go wrong and unexpected things are going to happen. Not being too optimized gives you a buffer to handle bad events. Having systems in place to take advantage of beneficial randomness gives better outcomes.
Clarity and trust - systems of trust reduce decision-making time and, along with ownership, provide the ability to surface better decisions from those closest to the issues.
Lack of overconfidence and rigidity - Plans change when they run up against the real world. Thinking that you can effortlessly accomplish an ambitious goal without problems is a recipe for disaster. At the same time, being too rigid in thinking about how something should be done prevents you from seeing alternative paths that provide a better way.
Incentives - having the right internal and external incentives will drive the behavior of the players of the system. For many projects in the current age, there are incentive structures that go unnoticed or appear as business as usual which make projects take longer than they should. Following what all players are incentivized by can tell you the likely dynamics and outcomes.
As a final thought, here's a quick story about the Empire State Building. Many different firms came in to bid for the construction of the building. A lot of groups came in with high flying proposals about how the work was an extension of projects that had been done before with "highly reliable estimates". Others talked about how new technologies would make construction faster. One group laid it straight to the owners. They told them that a building of this scale had never been built before and that new methods, which would be unknown before construction started, were going to need to be developed once construction was underway. Anyone that said they had a clear idea of how the Empire State Building could be built quickly, well, and on-time was yanking their chain. That group, that was honest about the task in front of them, ended up winning the bid and building one of the tallest buildings in the world, extremely fast.