Balaji Krishnamurthy
@_balaji_km
Followers
1K
Following
4K
Statuses
2K
In @Uber’s earnings slides and prepared remarks this quarter we went deeper on our AV strategy. We believe AV can open up a $1T+ TAM for Uber in the US alone -- but while AV tech is advancing, commercialization will happen much more slowly. Multiple elements still need to come together: ☑️a consistently super-human safety record - human-level safety is simply not good enough ☑️enabling regulations - which are still nascent in most markets ☑️a cost-effective, scaled hardware platform - most OEMs not able to produce at right cost or volumes ☑️excellent on-the-ground operations - Uber’s wheelhouse The fifth element is the most important: ☑️ high-utilization network that can manage variable demand with flexible supply ☀️🌛Ridesharing demand is highly variable, through the day, week, and year -- and no two cities are the same. In a typical large city, a fixed fleet designed to meet the weekly peak will have up to 95% of vehicles sit idle during the multiple weekly troughs. And a well-utilized network in March may not need ⅓ or more of its vehicles until October. 💰🚗Lastly, autonomy will unlock meaningful TAM expansion, but only when costs drop below human-driven rides, which are currently significantly cheaper on a per-mile basis.
43
47
308
There’s a lot more to it than putting a button on ___.
I think Ackman is wrong here on $UBER: if Google chooses to do so, it could put Waymo inside Google Maps (1 billion MAUs) and match riders to vehicles. Tesla is another threat. Either would obviate the need for Uber. All likely to happen a lot sooner than "many years".
1
0
40
UberX prices up marginally > "lower pricing environment" "To maximize demand, we remain committed to keeping prices as low as possible, passing through only the insurance cost increases to consumers; consequently, we expect UberX prices in the US to be up marginally in 2025."
Time to get off my tuchus
1
0
32
RT @dkhos: @BillAckman @Uber Love the ideas ... On multiple rides: we allow you to order an Uber for someone else now and allow multiple U…
0
2
0
RT @UberEats: In case you missed it 🏈 Greta Gerwig, Charli XCX, Martha Stewart, Sean Evans and Kevin Bacon star in a heartwarming film abou…
0
80
0
Good question - this is an evolving landscape. Over the long term, we expect fleet owners to emerge to own and manage the vehicles, while the other fixed assets (depots, charging infra, parking etc.) could be owned by REITs. We already have many fleet operators on Uber's network - nearly 50K fleets that manage over half a million drivers around the world. That said, in the near term, Uber might make some investments in fleets to kick-start the AV fleets ecosystem.
2
0
3
RT @TroyTeslike: I'm curious about what led to this incident with Tesla’s FSD V13. It was rainy and dark, and the Cybertruck didn’t detect…
0
36
0
RT @altcap: Never bet against @dkhos. Btw - the @BillAckman news dropped while Dara, Diller & I on stage together in Miami. At 33 M rides…
0
46
0
@LurkingRoss @Uber The hardest is that you have to solve all five together. Even if one of the elements doesn't come together, the puzzle falls apart.
3
0
9
Lots of respect for $MA and certainly a profile we can aspire for. Our goal is to drive growth for the long term, while delivering the highest FCF per share for investors over the long run. When our stock is cheap as it is today, we will be active and opportunistic, and reduce our share count. And regardless of the valuation, we will continue to improve the quality of our GAAP earnings. That means we will remain focused on SBC as a cost we have to actively manage, just like any other cost on the P&L.
3
0
4
Good analogy! In a market with these types of predictable, yet, extreme demand gyrations, a fixed pool of supply is capped to running near the trough levels of demand (to run the network efficiently). In other words, a 1P AV network can run efficiently only if they are build for 5-10% of the market, and paradoxically, dramatically lose utilization as they try to gain market share. That trade-off goes away with deployments on 3P networks.
@_balaji_km @Uber It’s very interesting — I found some strong parallels to the electricity market. In power, there are baseload power plants with high utilization rates and some supplementary peaker plants that only turn on for the peaks. AVs can do baseload but can they be the peakers?
6
1
23
RT @JonHartman14: @dkhos A reminder of the WSJ cover story on the eve of the IPO. Just keep executing.
0
1
0
Just leaving it here
👀 “A cost effective, scalable hardware platform”
0
0
6
Exactly right. There will be multiple winners with AV, it's just too large a TAM for a single winner to emerge.
There is a misplaced perception among some $TSLA bulls that the advent of unsupervised autonomy will somehow make $UBER obsolete and irrelevant when EV manufacturers can offer their own autonomous driving networks directly to consumers. Using this logic, one could argue $AMZN is irrelevant and uncompetitive to merchants who can sell directly to consumers. That logic is of course absurd, since AMZN’s 310M MAUs and vast distribution and fulfillment scale make it impossible for all but a few merchants to compete with AMZN. $TSLA ‘s ride hailing network will be insufficiently robust to compete with $UBER with 170M MAUs. Cheaper autonomous rides will be one option on the Uber app. If TSLA forbids its owners to join the UBER platform, other manufacturers will gladly fill the void. We are long both $TSLA and $UBER.
1
1
22
Good analysis here. Uber's Autonomous Future: From Network Effects to Platform Power
0
2
6
@treiner5 @bgurley Should say, super human safety is not there, for one. Also going to be challenged on most of the five pillars we articulated here
In @Uber’s earnings slides and prepared remarks this quarter we went deeper on our AV strategy. We believe AV can open up a $1T+ TAM for Uber in the US alone -- but while AV tech is advancing, commercialization will happen much more slowly. Multiple elements still need to come together: ☑️a consistently super-human safety record - human-level safety is simply not good enough ☑️enabling regulations - which are still nascent in most markets ☑️a cost-effective, scaled hardware platform - most OEMs not able to produce at right cost or volumes ☑️excellent on-the-ground operations - Uber’s wheelhouse The fifth element is the most important: ☑️ high-utilization network that can manage variable demand with flexible supply ☀️🌛Ridesharing demand is highly variable, through the day, week, and year -- and no two cities are the same. In a typical large city, a fixed fleet designed to meet the weekly peak will have up to 95% of vehicles sit idle during the multiple weekly troughs. And a well-utilized network in March may not need ⅓ or more of its vehicles until October. 💰🚗Lastly, autonomy will unlock meaningful TAM expansion, but only when costs drop below human-driven rides, which are currently significantly cheaper on a per-mile basis.
1
0
1