Why Uber’s Strategy Is Genius
Even When It Looks Stupid!
Uber is the kid everyone laughed at in high school…
who quietly spent a decade in the gym
and then showed up to the reunion looking like a Marvel character.
For years, the narrative was:
“They burn insane amounts of cash.”
“Unit economics will never work.”
“Regulators will kill them.”
And now?
Uber is a cash machine with real pricing power, 170M monthly users, and investors seriously arguing it’s still undervalued even after a 50%+ run-up this year.
There is only one claim:
Uber’s strategy looked stupid for a long time because it was designed for the endgame, not the early scorecard.
Let me break down what that actually means; strategy-style, not banker-speak.
1. Uber doesn’t run one business. It runs thousands of tiny local empires.
Everyone thinks of Uber as “one” marketplace.
Nope.
From a strategy lens, Uber is running hundreds (maybe thousands) of hyperlocal networks:
Drivers in New York don’t care about riders in London.
Riders in São Paulo don’t care about wait times in LA.
Each city is its own battlefield:
Its own driver supply.
Its own wait time.
Its own regulatory mess.
Its own demand curve.
What matters in each local market is simple:
Get enough drivers → keep wait time around 5–6 minutes → the network “clicks.”
Below 5 minutes?
You’re wasting capital adding extra cars no one needs.
Above 6–7 minutes?
People start opening another app.
The early “stupid” part of Uber’s strategy was spending ridiculous money to push every major city to that “sweet spot”; fast.
That looked dumb on an income statement.
But strategically?
They were buying local monopolies, city by city.
2. The London case: this is what winning looks like
Take London, one of Uber’s top 5 cities (those top 5 generate ~20% of Uber’s global gross bookings).
If you’re a typical Londoner:
You use Uber ~40 times a year.
You spend ~£650 annually on the app.
You can book: cars, food, groceries, trains, bikes, scooters, even flights.
And here’s the killer stat:
Uber has over 95% of private-hire drivers in London.
So imagine you’re a competitor like Bolt or FREE NOW.
You want to steal share:
You’re paying Google £0.50–£2 per ride just to acquire customers via Maps.
On a £17 ride, that’s 3–12% of the total fare… and up to ~45% of the platform’s own revenue cut.
Meanwhile, you still have to subsidise the fare to win the rider.
Then you need drivers:
New drivers are lured with weeks of zero commission.
Rough math in the article: acquiring a single driver can cost ~£2,600.
Poaching 20,000 drivers? That’s >£50M… in one city.
And even after all that?
Most of those drivers still multi-home (drive for both Uber + your app), and Uber still does 3x as many rides as you.
Translation: your spend becomes Uber’s liquidity.
Uber’s bet was:
“If we can get to dominant scale in each city, our competitors’ cost of acquisition becomes structurally stupid.”
London proves it worked.
3. How Uber killed multi-homing (on both sides)
In the early days, everyone multi-homed:
Riders: 2–3 apps on their phone.
Drivers: online in multiple apps, chasing surge.
That’s death for margins.
Uber’s “genius but boring” moves:
a) Stuff the app with everything
Not just rides. In many cities you can:
Order food (Uber Eats)
Groceries
Retail items
Trains, buses, boats, flights
Bikes & scooters
Multi-product users:
Take 2–2.5x more trips than single-product users.
Generate 3x more gross bookings.
Are ~25–50% cheaper to acquire than via paid channels.
Suddenly, choosing a competitor isn’t just “Is this ride cheaper?”
It’s “Am I ditching my entire habit stack across rides + food + grocery?”
b) Membership: Uber One as the glue
Uber One (their membership):
Went from 5M members → 30M in a few years.
That’s ~19% of total users.
Those members drive ~35% of total Mobility + Delivery gross bookings.
And around 60% of Delivery gross bookings.
Pay a small monthly fee, get:
$0 delivery fees,
Service fee discounts,
Mobility perks.
If you order >3 meals/month, it’s a no-brainer.
For Uber, it’s:
High-margin subscription revenue (about $5/month per member).
Higher frequency.
Lower churn.
Membership is how you turn “I compare prices across apps” into “I just use Uber, it’s already paid for.”
c) Pricing games that make comparisons hard
Uber added things like dynamic pricing and upfront pricing:
Better match supply/demand.
Quietly raise average fares.
Increase take-rate (the % of the fare they keep).
Lyft copied similar pricing tactics.
Now in the US they’re basically a cozy duopoly, controlling almost 100% of the rideshare market and avoiding all-out price wars.
When both players move from “growth at all costs” → “margin expansion,” that’s when the business goes from Twitter-thread joke → monster cash engine.
4. From cash bonfire → cash machine
For years, Uber looked like the poster child of “VC-subsidised insanity.”
Now?
8M+ drivers & couriers
1M+ merchants
170M monthly active users
Advertising already >1.5% of gross bookings (high margin).
Promotional costs per user are down +50% in the last 2 years.
Group EBITDA margin around mid-teens and climbing.
At today’s price (~$90 at time of writing), the market is valuing Uber as if:
It never meaningfully grows beyond its current scale, or
Its pricing power + membership + ads are already fully priced in.
Their view: that’s wrong.
With:
Active users still growing ~8–10% annually,
Trips per user growing,
Price per trip trending up,
Take-rates rising in both Mobility and Delivery,
…they think Uber can grow free cash flow massively over the next 3 years and is worth around $143/share; ~60% upside from that $90 mark.
That’s not some early-stage dream.
That’s just the math of:
more users × more trips/user × higher revenue per trip
higher-margin layers (ads + membership)
– lower marketing per user
5. “But what about the risks?”
The big ones:
Regulation – Uber’s “ask forgiveness, not permission” playbook got them kicked around a lot (see: London license drama). But now regulation is mostly a barrier to new entrants. It’s ugly, but it locks in the incumbent.
Insurance – Up to a third of the fare can be insurance. Rising premiums hurt affordability… but again, that makes life harder for under-scale competitors even more.
Recession – Less demand from riders, more supply from unemployed drivers. Net effect unclear, but Uber may become more competitive vs smaller players.
Autonomous vehicles – Could OEMs (Waymo, Tesla, etc.) disintermediate Uber? Maybe. Uber’s hedging with multiple partnerships instead of building its own AV stack now. Management doesn’t see AV as a huge % of gross bookings “for many years.”
Freight – Still small, unprofitable, and basically treated as a free option in the valuation.
None of these are trivial.
But the pattern is clear: in messy, regulated, capital-intensive spaces, scale is the moat.
Uber already did the stupid-hard, capital-inefficient part.
6. What you can steal from Uber’s “stupid” genius
You’re probably not building a global rideshare app.
But the strategic lessons are portable:
Play for the endgame, not the early scoreboard
Looking unprofitable in year 3 is fine if you’re building something that’s unassailable in year 10.Design for local dominance, then scale the playbook
Uber wins city by city. What’s the “city” in your market? A niche, a vertical, a geography?Kill multi-homing with bundles + membership
Don’t just be “the cheapest.” Be the default. That’s what multi-product + membership did for Uber.Use regulation & complexity as a moat
If your space is a nightmare to understand and navigate, you might be in a great business if you’re willing to suffer longer than others.Don’t be embarrassed by the ugly years
Every “overnight” monopoly spent years looking stupid to people reading only the P&L.
Uber looked insane…
until suddenly, it didn’t.
That’s the whole game.
– Houman,
Full Stack Capitalist