He spent the last 8 years building a new company in stealth (for a long time, employees weren’t even allowed to list it on their LinkedIn).
In March he shared the big picture vision for his new company, Atoms and then this week announced a $1.7 billion round led by a16z.
I was working at Uber in 2017 when he was ousted as CEO. Generally speaking, the world got the story wrong on him then, so I love to see the comeback now.
I watched his interview with Ben Horowitz and Erik Torenberg from a16z in full and while here were a lot of great takeaways, I want to focus on one in particular…
Btw here’s the full interview (it’s worth a watch):
And Then You Say LFG
Being a founder is largely about complexity management.
There’s simply no way to predict either all of the ways things are going to go wrong, nor all of the creative ideas you’re going to have along the way to make things go better.
For example, when Jeff Bezos started Amazon do you think he knew one day they’d go from an online bookstore to running a giant cloud computing platform that drives 20% of their revenue?
No, of course not, because it doesn’t make sense at face value.
Even in Elon Musk’s case, with his Tesla Master Plan, do you think he foresaw that the original Tesla Roadster would lead to… humanoid robots?
Probably not.
But in both of these cases they and their teams took on a new and audacious challenge to grow the business in a direction that the world was already heading because they had already built up what happened to be structural advantages for these new markets with their existing businesses, and wouldn’t need to break their focus on the existing businesses to do it.
Travis says all founders should be aware of this and actively cultivate it. He calls it “the meta problem.”
In the interview with a16z, Travis shared his framework for it.
He defines it as “the derivative of your problem creation factor must be less than or equal to the derivative of your problem solving factor.”
He’s saying that starting to pursue AWS or Optimus is a smart decision only if the added complexity from managing those new “problems” is within your capacity to manage them.
If that isn’t true, you’re “underwater” and need to stop problem creation right away until it changes.
He notes that when he was CEO of Uber DoorDash had 5% of the food delivery market and Uber Eats was much larger, and the autonomous driving division that today would have made the company much larger than it is was the second best in the world to Waymo “but catching up.”
Who’s to say for sure if he’s right, but the framework holds.
Ben Horowitz even notes in their conversation that Uber would be “a much larger and more significant” company today if Travis hadn’t been ousted.
Either way, the lesson here is that when making a big, multi-year bet there’s a way to identify if it’s worth going in on or not:
Paint the vision
Go to the most thoughtful and knowledgeable you know both within and outside of your company. Tell them your vision for the market you’re attaching 15-20 years in the future.
For example, "In 20 years, will high-quality meals be made and delivered by robots at near-grocery prices?”
If you can articulate a simple vision in a way that no credible person easily disputes, move forward.
Compress the Timeline
Twenty years is too long for most founders and investors.
So, if there’s a big vision, try to push the limits of how soon it could happen.
Ask the harder questions of yourself first:
What about 10 years?
7 years?
5?
If you find a seemingly viable path to make it happen at that point, go back to those same people and present it to them. See what they say.
This is when you as the founder are bending reality towards now. You’re not just following what the world thinks will happen, you’re exerting your will and vision onto the world.
You’re changing who pulls it forward from “someone” to yourself.
When & Who
If you’re getting intense debate from the smart people around the fundamental premise, you’re not ready.
You haven’t been able to find the truth yet, or you haven’t found a way to bend it forward in time in a believable enough way.
But if the only remaining questions are debates on when someone will do it and who that will be, that’s the moment you know you have something.
Check Your Management Capacity
Use Travis’s complexity management framework to check: are you the person to pull this off?
In Travis’s and Jeff’s and Elon’s cases, they all have trended towards multi-product, multi-business-line, multi-revenue stream work. Having that level of complexity management is extremely rare at the level they’re doing it.
Obviously, you can’t predict all of the complexities that will come up, but just in terms of the core problem you’re going after be honest with yourself. Can you bend reality in this way?
Make that determination for yourself…
And then you say LFG.