Next Sunday, June 22, Tesla’s Robotaxi service will be open to the public “tentatively.” That means about 10 Tesla-owned Model Y’s will be part of the launch inside an Austin geofenced area during the first week, about five years behind Elon’s initial rollout timetable. Adding some sizzle to the autonomy topic, the company will have the first Tesla that drives itself from the factory end of line all the way to a customer house on June 28 (both of which will be in Austin).

While the initial launch will be small, it will be closely watched because it’s the company’s second major step (behind FSD beta) into a higher margin, autonomy-first business model. Musk told CNBC a few weeks ago that the fleet could grow to about 1,000 vehicles “within a few months” if early results meet expectations.

Musk has indicated that Tesla’s most likely long-term go-to-market approach will center on an asset-light model, where consumers allow their cars to operate in the fleet when not in use. Additionally, I expect small businesses to emerge that manage micro fleets. Market share will likely resemble today’s Uber and Lyft split, around 70% to 30%.

The asset light approach has three primary benefits. First, it scales faster given it can leverage the 7.7m vehicles in the US by mid-year 2025 that will be able to contribute to the fleet. Second, it’s financially less risky given the capital risk is on the owners and Tesla takes a 30% cut of the revenue. Third, the business should have higher margins. The downside to the asset light approach is that total profit dollars will be less compared to the owned and operated asset heavy approach.

By 2030, if Tesla offered rides at a 10% discount to Uber and Waymo and they can:

• Capture 30% of the US ride sharing market, it would add about $8.6B to operating income, or increase overall operating income by 32%.
• Capture 70% of the US ride sharing market, it would add about $20B to operating income, or increase overall operating income by 74%.

The sensitivity above assumes the average price of a ride remains unchanged over the next five years at $22 ($20 factoring in Tesla’s market share gain strategy of offering a 10% discount). A more aggressive scenario assumes the price of a ride increases 5% per year over the next five years. Under that scenario:

By 2030, if Tesla offered rides at a 10% discount to Uber and Waymo and they can:

• Capture 30% of the US ride sharing market, it would add about $10.8B to operating income, or increase overall operating income by 40%.
• Capture 70% of the US ride sharing market, it would add about $25.2B to operating income, or increase overall operating income by 94%.

Key market assumptions:

• There were about 4B ride share rides in the US in 2024. The “rides” growth rate for Uber and Lyft in 2024 was around 25%, and it is expected to decline to about 12% for 2025. We believe a reasonable average annual rides growth rate through the end of the decade is 5%.
• That means by 2030 there will be about 5.3B rides per year in the US. Our sense is this is a conservative number, given that as autonomy drives down the average cost per ride, we should see growth in the number of rides accelerating.
• Giving a 10% discount to the current average price per ride would be compelling to riders, going from today’s $22 per ride to $20.
• The combination of increased rides and a lower price per ride results in the addressable market growing from $88B today to $106B in 2030.
• Our more aggressive sensitivity assumes the average price per ride increases by 5% per year.

Tesla’s less likely approach is the asset heavy model where Tesla owns the fleet. In that scenario, the market share will also likely mirror what we see today with Uber and Lyft, splitting it 70% to 30%.

The asset heavy approach has one primary benefit: it generates more absolute profit dollars. The downside is it scales slower, given Tesla would need to manufacture somewhere between 550k and 1.3 million vehicles to service the 30–70% market share. Additionally, it’s a lower margin business.

By 2030, if Tesla offered rides at a 10% discount to Uber and Waymo and they can:

• Capture 30% of the US ride sharing market, it would add about $14.3B to operating income, or increase overall operating income by 53%.
• Capture 70% of the US ride sharing market, it would add about $33.8B to operating income, or increase overall operating income by 125%.

A more aggressive outcome: the sensitivity above assumes the average price of a ride remains unchanged in five years at $22 ($20 factoring in Tesla’s market share gain strategy of offering it at a 10% discount). A more aggressive scenario assumes the price of a ride increases 5% a year over the next five years. Under that scenario:

By 2030, if Tesla offered rides at a 10% discount to Uber and Waymo and they can:

• Capture 30% of the US ride sharing market, it would add about $18.3B to operating income, or increase overall operating income by 68%.
• Capture 70% of the US ride sharing market, it would add about $43.1B to operating income, or increase overall operating income by 160%.

The need to sell new vehicles for the fleet depends on how many of the 7.7 million existing Tesla owners choose to opt in. If many do, there won’t be much demand for consumers or small businesses to buy the Cybercab. But if few opt in, demand for new Cybercabs will be higher.

On Tesla’s March earnings call, Musk reiterated that the company expects to ramp production of the Cybercab sometime in 2026, or in about a year and a half from now. Additionally, Musk is “aiming for at least 2 million units a year of Cybercab. That will be in more than one factory, but I think it’s at least 2 million units a year, maybe 4 million ultimately.” I see the need for 2–4 million Cybercabs as unlikely. I estimate the US can be served by 550k to 1.3m vehicles, and some of the demand will be fulfilled by existing owners. That leads me to believe the annual Cybercab run rate will be between 200k and 400k a year by 2030. Growing current Tesla deliveries at an average of 15% per year through 2032 implies that Cybercab (at 300k annual deliveries) would account for 5% of deliveries in 2030.

These are exciting times with the progress that both Waymo and Tesla have made over the past year. That said, it’s going to take at least 4 years before ridesharing is dominated by autonomy.

Below is our best guess of what the conservative ramp will look like in terms of the percentage of rides that are autonomous:

Conservative ramp:

2025 – 0%. Waymo will hit about 13 million rides this year, or about 0.003% of the market
2026 – 5%
2027 – 8%
2028 – 10%
2029 – 20%
2030 – 30%

Aggressive ramp:

2025 – 0%. Waymo will hit about 13 million rides this year, or about 0.003% of the market
2026 – 10%
2027 – 18%
2028 – 30%
2029 – 50%
2030 – 70%

The biggest factor in the speed of the ramp is getting the green light from federal and state governments to launch the services. The second biggest factor is the pace at which existing Tesla owners opt into the fleet.

Government: As for the federal and state governments getting behind this, time will tell. If Tesla’s Robotaxi service is accident free in 2025 and Waymo continues to have a good record, the odds are favorable that the trend builds momentum with lawmakers. If the safety factor stumbles—which I put in the camp of more than 10 accidents this year—the red tape will make this more difficult.

Tesla owners: My sense is only a small percentage of the 7.7 million eligible Teslas will opt in, and that’s OK because by my math only 17% (1.3 million vehicles) need to opt in to service 70% of the market. The pace of Tesla owner buy-in to the fleet will be dependent on the economics.

Some quick math for perspective:

• Car “works” for an average of 4 hours a day
• Earning $20 a ride, two rides per hour
• Car works 3 days a week
• Pays Tesla a 30% take rate
• Pays $1.2k per year for FSD
• Pays $500 per month for insurance, charging, and maintenance
• Yields annual before-tax income of $10k

On paper, that’s real money. In practice, the cars will get beat up, need to be cleaned up, which means it’s unlikely that more than 20% will ever opt into the fleet.

Small business fleets: The network will also get additional supply from businesses that operate a fleet of 5–10 cars. Here’s a look at their economics:

• Fleet of 5 cars
• Earning $20 a ride, two rides per hour
• Each car operates 10 hours a day
• Fleet operates 5 days a week
• Pays Tesla a 30% take rate
• Pays $1.2k per car per year for FSD
• Pays $1.4k per car per month for lease, insurance, charging, and maintenance
• Yields annual before-tax income of $275k