$Uber(UBER)$ I met with Travis Kalanick on a panel at UCLA back in 2016.
He was still the CEO of Uber at the time, and he was answering questions ranging from entrepreneurship, corporate governance, how to build, how to scale, etc..
What stuck with me from that panel is how the UBER business model evolved over time. It didn’t happen. It happened because of a critical insight—an insight they gained after they started building the company.
The spark for Uber came in December 2008, when Garrett Camp and Travis Kalanick were frozen out of Parisian taxis after the LeWeb conference.
Don’t imagine two struggling engineers; at this point, both of them were already rich. Kalanick had sold Red Swoosh to Akamai Technologies for $19 million, while Camp sold StumbleUpon to eBay for $75 million a year ago.
When they somehow got themselves in a cab, the same thought was going through their entrepreneurial cortex, “Getting a cab shouldn’t be this hard.”
The reason I said you shouldn’t imagine two struggling engineers is that what they later built was a solution to their own problem.
These were two wealthy entrepreneurs. Kalanick explained that they weren’t seeking cheap, accessible mobility; they were looking for something fast, comfy, and premium.
They did exactly this. The first UBER they built wasn’t a basic taxi app—it was a premium driver service.
They worked with licensed limo drivers, and when users requested a cab, they would be picked up with a black, luxurious sedan, of course, at premium prices.
They named it Uber Cab and launched the beta in 2009. This was its website:
You see what this is? It has nothing to do with the current Uber model we now know.
Kalanick made this very clear in the panel. His sentences stuck with me: “We weren’t the geniuses who envisioned Uber’s today, we simply created a premium taxi business that you could use through a website or app. We solved our own problem.”
How did it morph into the Uber we know today? It was a contingent path.
Their early hires were Google data engineers because they were using Google APIs for mapping and calculating the estimated time of arrival.
That team generated a crucial insight—when the estimated time of arrival (ETA) was below 5 minutes, conversions surged.
From there on, the challenge was obvious: How to reduce ETA?
At any given moment, there were more prospective riders than drivers. This was the main factor leading to a longer ETA. How to solve this? Remove the lid that constrains the supply—this was the crucial insight.
At around the same time Lyft started experimenting with peer-to-peer riding services, and California’s Public Utilities Commission signalled it would treat “transportation network companies” differently from limos, opening a legal path for private‑vehicle drivers.
And Uber Cab made the Pivot to UBerX.
They incorporated a peer-to-peer platform in their application, allowing third-party drivers to respond to demand. Growth exploded.
In his own words, this was when Kalanick understood that “distribution is the king.”
UBER stock price has been depressed lately because of worries that the robo-taxis will disrupt its ride-hailing business, relying on human drivers.
I bet it won’t. Further, I bet it’ll make Uber even richer.
I believe those who argue otherwise largely misinterpret UBER’s business, they ignore a crucial fact—distribution is the king, and UBER has it.
Today, I’ll explain why UBER is one of my high-conviction picks for the next decade.
Let’s cut the BS and get started!
What you're going to read:
1. Understanding The Business
2. Competitive Analysis
3. Investment Thesis
4. Fundamental Analysis
5. Valuation
6. Conclusion
What Kalanick referred to as “distribution is the king” is actually a profound insight into platform economics. UBER isn't just a ride-hailing app — it's one of the most sophisticated business ecosystems ever built.
This is what most people don’t get about it, and most of those who get it don’t really understand the huge implications of this in terms of business economics.
Let’s dig here a bit.
The pivot from UberCab to UberX represented more than just a business model shift. It unlocked what economists call exponential network density — a critical factor that makes or breaks marketplace businesses.
Network density measures how many potential matches exist within a given geographic area. When UBER allowed anyone with a qualifying car to become a driver, it dramatically increased this density.
At its core, UBER's platform executes a complex optimization problem across multiple dimensions:
Geographic matching — Uses GPS data to pair riders with nearby drivers
Dynamic pricing — Adjusts prices in real-time based on supply-demand imbalances
Dispatch efficiency — Calculates optimal routes considering traffic patterns
Multi-homing — Enables drivers to serve both ride requests and deliveries
Batching — Groups multiple requests to maximize driver utilization
The technical implementation is staggering. UBER's systems process over 100 million daily requests, each requiring calculations across billions of possible combinations in milliseconds.
Every ride teaches the algorithm something new — traffic patterns, rider preferences, driver behavior. This creates a data flywheel that compounds UBER's advantage. With each additional transaction, the platform becomes incrementally more efficient.
This increasing efficiency results in Uber keeping more of the customer payments. Its take rate has steadily increased from ~20% to ~27% in mature markets.
This revenue model scales exceptionally well because:
Fixed costs are largely decoupled from transaction volume.
Technology investments benefit the entire network.
New verticals leverage the same infrastructure.
This platform extensibility enabled UBER to build additional businesses on top of its core infrastructure — becoming an ecosystem of integrated businesses.
UBER Eats transformed restaurant delivery using the same matching algorithms. It connects restaurants, drivers, and hungry customers. Launched in 2014, it now processes $70 billion in annual bookings across 11,000+ cities.
UBER Freight digitized the inefficient $800 billion trucking industry. Traditional brokers use phone calls and paperwork. Freight offers instant quotes, real-time tracking, and automated payments.
Why is an ecosystem a huge advantage? Because it creates cross-platform efficiencies.
For example, a driver might transport a passenger to work, deliver lunch via Eats in the afternoon, then transport packages before evening rush hour. This cross-utilization creates operational leverage that no competitor currently matches.
This advantage translates directly into pricing power. As UBER continues to optimize its matching algorithms and expand density across verticals, it strengthens its position as the default transportation platform.
This is what Uber really is. It’s an ecosystem.
But how durable is this model?
Is it also durable against what’s coming—autonomous taxis?
I think it is.
Competitive Analysis
Let me be straightforward—UBER defies the traditional norms of competition.
This is why I love it.
Traditionally, competitive advantage is thought to stem from two sources: Differentiation and lower costs.
The godfather of the competitive strategy, Michael Porter, institutionalized the concept in his great book “Competitive Strategy” published in 1980.
It’s really simple: A Product or service can either be generic or specific.
The model relies on two assumptions:
If the product is a generic one, people will buy the cheaper alternative.
If it’s specifically tailored for one job, people will buy the one that does it best.
If you go to a business school today, this is still the model taught. Because it works, at least most of the time. At the time of this model, they thought people would buy either because it’s cheaper or because it’s the best. They didn’t think people could buy or use something because other people do.
Indeed, the concept was first laid out in 1985, 5 years after Porter’s book, by Carl Shapiro and Michael Katz in their seminal paper “Network Externalities, Competition, and Compatibility.”
What’s special about these guys? They are competition economists from UC Berkeley.
Being from Berkeley, they were well immersed in the competition of technology companies surrounding them. Looking at Microsoft, they noticed a phenomenon—as the number of people using the Microsoft operating system increased, more people were attracted to it.
They observed that as more people used the system, it became more valuable simply because of compatibility factors, and more users got attracted to be compatible.
They called this “network externalities.” Today, it’s usually called direct network effects.
Social networks like Instagram and Facebook are the best examples. The bigger they get, the more valuable they become for users because you can find more friends there, create larger audiences, etc..
Direct (basic) network effects
Strong direct network effects make it extremely hard for competitors to disrupt the incumbent, yet they aren’t the strongest form of network effects.
In the late 1990s, we understood that the internet enabled yet another new business model: Two-sided marketplaces.
Think about Amazon. It’s a marketplace with two sides, buyers and sellers.
More buyers make it more valuable for sellers and attract sellers.
More sellers make it more valuable for buyers because sellers engage in price competition, and low prices attract more buyers.
This keeps working infinitely, creating a virtuous cycle.
After observing companies like Amazon, economists Jean-Charles Rochet and Jean Tirole formally defined two-sided marketplaces in their landmark 2003 paper “Platform Competition in Two-Sided Markets”.
In their framework, Uber exhibits classic indirect network effects: the value to riders increases with more drivers participating, and vice versa.
This flywheel creates what I call a compounding marketplace advantage. Each incremental transaction strengthens Uber's position relative to competitors
Let me explain:
When Uber enters a new market, initial ETAs might be 15+ minutes. As they add drivers, ETAs drop to sub-5 minutes, where conversion rates soar. Once a market reaches critical density, something remarkable happens — wait times stabilize at 2-3 minutes while driver utilization simultaneously improves.
This is mathematically impossible for subscale competitors to match. If a competitor has 1/10th the drivers, they cannot achieve comparable ETAs while maintaining the same utilization rates. And they’ll likely keep having fewer drivers because Uber’s indirect network effects make new drivers choose UBER rather than competitors.
So, why did I explain all this? Everybody has already agreed on Uber’s supremacy over rivals. The real threat is robo-taxis. Right?
The same dynamics apply to robo-taxis, too. Think of it like you are just changing the driver. Instead of humans driving the car and making the money, it’s now AI models driving the cars, and companies behind them are making the money.
What does this change in Uber’s position mean? I think, nothing.
Remember, the main success of Uber is that it successfully aggregated enough of a commoditized service, in this case, human-driven taxis. The same aggregation problem exists if the new service is also a commodity. I think it’s.
Earlier, much of the thesis about robo-taxis was that Tesla would dominate it. However, Waymo has already launched, it’s operational in several cities, completing more than 250,000 paid drives a week, and Tesla is still not around.
On top of that, Nvidia is developing a generalized infrastructure for autonomous driving and a specialized infrastructure for General Motors. Nebiu’s AVride has also completed more than 10 million miles on public roads. It’s coming too.
If this technology had been reserved for one early mover and it’ll have long enough time alone in the market to reach a critical mass before others could launch, I would say it’s a threat to Uber. But in the current shape, it’s already clear that the tech will be commoditized, and we’ll have many providers.
Where the product or service is commoditized, distribution is the king.
If these companies try to scale globally at the same time, that’ll just be a race to the bottom. They’ll each struggle pretty much with the same things, creating a huge deadweight loss in the industry and delaying profitability for years. Plus, some of these services will face the risk of staying below the efficient scale.
What I think will happen instead is that most of those providers won’t want to assume the risk of staying below the efficient scale. Instead, they’ll partner with Uber to leverage its distribution. We already see this:
What I think will eventually happen is that Uber will evolve into the Amazon of mobility.
Just as Amazon connects buyers with third-party sellers of commoditized products, Uber will connect riders with commoditized autonomous transportation services. Uber will generate revenue through platform fees, dispatch optimization, and customer management—all while maintaining its human driver network in jurisdictions where autonomous vehicles aren't yet permitted.
This is an amazingly strong competitive position to be in at this stage.
In short, robo-taxis won’t disrupt Uber, they’ll make it bigger, faster, and more profitable.
The king of distribution becomes the emperor of commoditized supply.
This isn't just an advantage; it's an economic inevitability in markets defined by indirect network effects, platform economies, and regulatory complexity.
Comments