Tesla Network - for profit ride sharing managed by Tesla

[Ludus]

think that is a entirely fair assessment. I love tesla and have encouraged as many people as I can to preorder model 3s. But I dont expect to ever use mine for personal use. It seems like a waste considering the potential margins, and as you point out, probably not going to last more than ~5years. For example, estimating Teslas production: 500,000 yr1 + 1mil yr2 + 1.5 mil yr3 + 2mil yr4 + 2 mil yr 4 = 7 million self driving taxis worldwide by 2022. Uber has like 300,000 drivers that can only work 40 hr weeks. So effectively tesla network would be 100 times bigger than Uber is today. That would be a pretty saturated supply of self driving cars. Granted with the exception of national boarders they could self organize and make sure the distribution of taxis was equal all over if permitted, so you'd have service in the middle of Kansas potentially. I imagine at that point tesla would be paying you an hourly rate rather than people that actually got in your car, just to maintain market share everywhere they could. But at some point soon after that Im betting cars will be rolling off the factory line either being a tesla owned taxi (because theyd have a ton of money by then) or be a personal vehicle but models produced after x date wouldnt be allowed to drive on TN.

When Tesla starts producing Model 3, it will be distributed to customers all over the world and in most of the world TN L5 autonomous fleets won't be legal. My guess is that Florida will see the first TN fleet in 2017-2018, but it won't be legal in any other state then.

The TN fleets would be in concentrated service areas that provide saturation so most of the vehicles will be company owned. Tesla would allow owners who happen to live in service areas (say in Tampa Bay Area) add their cars to the fleet, but they have no reason to let third parties buy up Teslas for that purpose any more than they would allow Teslas to operate for Uber or Lyft.

Over time there would be enough owners in service areas to matter and Tesla could accommodate their participating in TN easily.

I think TN will accommodate more sorts of uses than Uber. For example a sort of friends and family shared use that doesn't charge and only allows a very limited group of users. That might be available as soon as it's legal in a jurisdiction, while Uber like service to the general public would have to wait for Tesla to saturate a service area and roll it out.

 

[Green Pete]

When Tesla starts producing Model 3, it will be distributed to customers all over the world and in most of the world TN L5 autonomous fleets won't be legal. My guess is that Florida will see the first TN fleet in 2017-2018, but it won't be legal in any other state then.

The TN fleets would be in concentrated service areas that provide saturation so most of the vehicles will be company owned. Tesla would allow owners who happen to live in service areas (say in Tampa Bay Area) add their cars to the fleet, but they have no reason to let third parties buy up Teslas for that purpose any more than they would allow Teslas to operate for Uber or Lyft.

Over time there would be enough owners in service areas to matter and Tesla could accommodate their participating in TN easily.

I think TN will accommodate more sorts of uses than Uber. For example a sort of friends and family shared use that doesn't charge and only allows a very limited group of users. That might be available as soon as it's legal in a jurisdiction, while Uber like service to the general public would have to wait for Tesla to saturate a service area and roll it out.

In general I'd agree. But if a self driving car can work 120 hour weeks rather than 40 that means 1 TN vehicle = >3ubers (uber drivers on average drive less than 40 a week). Uber was in most major cities in the us when they had 300,000 drivers globally. But let's say all 300,000 were in the Us. For equal the coverage of Uber it would require ~100,000 cars driving on a pretty heavy schedule. Or 150,000 driving at 80 hour weeks to match uber. There are 400,000+ preorders. So as far as market saturation that's not likely to take long at all, I see it possible for 150,000 cars to hit the road by January 2018.

Now, getting all states to approve it could take longer.

What's your reason for Florida? I would think California would be first. I am sure the Tesla team is working closely with California regulators as that's there home base.

 

[Ludus]

From the transcript

"Sometimes, it's been characterized as Tesla versus Uber or Lyft or something like that. It's not Tesla versus Uber; it's the people versus
Uber."

Here is the full section:

Charlie Anderson:
"Thank you. I'll ask just one question. There was a reference to the Tesla network and the ability to buy self-driving today. So I wonder, Elon, if you could talk maybe philosophically about how you're viewing Tesla network. Is it something that will generate income for Tesla? Does it help develop future products, etcetera, at a reasonable gross margin? Or is it something that you'll use more for market share gain, help people offset the price of the car long-term? Thanks"

Elon Musk:
"All right. Sorry. Just talking internally for a second there. I think it's a bit of both, really. This would be something that would be a significant offset on the cost of ownership of a car and then a revenue generator for Tesla as well. Obviously, the majority of the economics would go to the owner of the car. Sometimes, it's been characterized as Tesla versus Uber or Lyft or something like that. It's not Tesla versus Uber; it's the people versus Uber."

You can read that several different ways. He's not denying that it would be a profit center for Tesla. I think that phase comes first when Tesla needs it most to generate capital for building factories. It's pretty obvious that it will only be legal gradually place by place even though Tesla will be shipping cars everywhere so the owner supplied Tesla Network is a future thing as the density of owners gets higher in some places. That option definitely supports owners and encourages sales. It's just that the owner supplied Network isn't a major factor for quite a few years, TN starts mostly company owned.

The "people vs Uber" highlights the fundamental problem Uber has with it's business model that Tesla doesn't have. Uber is trying to screw and replace the drivers who are the foundation of it's business. Tesla is happy to integrate it's owners and share the revenue and has no conflict.

 

[GoTslaGo]

In general I'd agree. But if a self driving car can work 120 hour weeks rather than 40 that means 1 TN vehicle = >3ubers (uber drivers on average drive less than 40 a week). Uber was in most major cities in the us when they had 300,000 drivers globally. But let's say all 300,000 were in the Us. For equal the coverage of Uber it would require ~100,000 cars driving on a pretty heavy schedule. Or 150,000 driving at 80 hour weeks to match uber. There are 400,000+ preorders. So as far as market saturation that's not likely to take long at all, I see it possible for 150,000 cars to hit the road by January 2018.

Now, getting all states to approve it could take longer.

What's your reason for Florida? I would think California would be first. I am sure the Tesla team is working closely with California regulators as that's there home base.

I wonder if this will make "surge" pricing not needed. My understanding is that Uber uses this to bring out more drivers to provide capacity during off-peak times and odd hours. The riders obviously pay for it. With Autonomous vehicles there's no impact to the driver's lifestyle/sleep, etc... This may increase TN demand since they may not need to charge surge pricing, especially off hours. Now if Uber is also successful with autonomous...

 

[Ludus]

In general I'd agree. But if a self driving car can work 120 hour weeks rather than 40 that means 1 TN vehicle = >3ubers (uber drivers on average drive less than 40 a week). Uber was in most major cities in the us when they had 300,000 drivers globally. But let's say all 300,000 were in the Us. For equal the coverage of Uber it would require ~100,000 cars driving on a pretty heavy schedule. Or 150,000 driving at 80 hour weeks to match uber. There are 400,000+ preorders. So as far as market saturation that's not likely to take long at all, I see it possible for 150,000 cars to hit the road by January 2018.

Now, getting all states to approve it could take longer.

What's your reason for Florida? I would think California would be first. I am sure the Tesla team is working closely with California regulators as that's there home base.

You're right from the high level view that Tesla could in raw numbers equal Uber very quickly, it's the details in the weeds that interfere.

I think the company TN fleets will be on the street 24/7 so 168 hours a week. Charging just takes some snake Superchargers scattered around the service area. The recent move to make Superchargers pay as you go facilitates this and I think the snake versions have to be in place at least selectively for that cross country L5 trip in 2017.

Florida is the only place in the US and possibly the world at the moment that has legalized commercial operation of autonomous vehicles. It's there in state law. Michigan, Nevada, Cali have limited permission for R&D with a driver ready to take over. Most states don't formally permit or prohibit it so you could get away with passing through but not a high profile service. This might change in time but Tesla is trying to bring this out a few years ahead of everybody else so it might not.

Tesla has no plans to change it's sales model so it will be sold like a conventional vehicle. A pretty large fraction of owners won't be interested in loaning them to TN and an even bigger fraction won't live in a viable service area. The fraction that live in the right place and want to loan out their vehicle may not want to do so when it's needed.

So from the ground view, Tesla opens a service area in Tampa Bay and needs a couple thousand vehicles 24/7 to saturate before it can switch on App service. It brings in it's own and starts the service 100% company. It now gets the data from actual operations there and knows better what it needs and can adjust. It opens Tampa TN to Tesla owners. How many are there at the times required? Tesla will let owners displace company vehicles if they can cover the demand because it can just redeploy the company cars to another new service area. The question is how quickly does this happen? I think it takes years to get to 50% in most service areas. Being able to loan your car to TN would increase sales in an area.

I'm assuming Tesla devotes some fraction of it's production to company owned TN vehicles. The higher the fraction the faster it could displace Uber in the most lucrative markets.

If it's not constrained by regulation, they'd want to go into NYC, SF, LA, Chicago, London, Paris etc. and take over those high value markets. For NYC that might take at least 5000 vehicles but could probably use 20,000. When the NYC area has say 200,000 Tesla owners they might provide most of the required vehicles with minimal need for company owned.

 

[Ludus]

I wonder if this will make "surge" pricing not needed. My understanding is that Uber uses this to bring out more drivers to provide capacity during off-peak times and odd hours. The riders obviously pay for it. With Autonomous vehicles there's no impact to the driver's lifestyle/sleep, etc... This may increase TN demand since they may not need to charge surge pricing, especially off hours. Now if Uber is also successful with autonomous...

That depends on how you think of the fleet ownership. If TN is mostly company owned then prices are a ceiling not a floor. There is no surge pricing. Instead there are discounts and give aways. A major reason Uber burns cash is trying to dominate service areas by paying drivers so it can discount. Maybe something like TN Prime with deep discounts during off hours.

As TN service areas are more supplied by owners it might come to be useful to have some form of surge payment to them to encourage them to loan their vehicle during peak demand periods,

 

[adiggs]

I wonder if this will make "surge" pricing not needed. My understanding is that Uber uses this to bring out more drivers to provide capacity during off-peak times and odd hours. The riders obviously pay for it. With Autonomous vehicles there's no impact to the driver's lifestyle/sleep, etc... This may increase TN demand since they may not need to charge surge pricing, especially off hours. Now if Uber is also successful with autonomous...

My understanding of surge pricing is for opposite reasons, but also possibly completely compatible. Namely, that surge pricing arises in areas where there is a surge in demand. The mental model I have is the end of a professional basketball game, or the start, where people are using Uber / Lyft to get a ride to and from the game. The surge pricing is designed to bring an increase in supply to an area that lacks adequate supply for the demand.

The latter statement makes the high demand and low supply periods equivalent for the purposes of surge pricing - in both cases there is a mismatch between demand and supply, with demand being higher than supply.

 

[Green Pete]

My understanding of surge pricing is for opposite reasons, but also possibly completely compatible. Namely, that surge pricing arises in areas where there is a surge in demand. The mental model I have is the end of a professional basketball game, or the start, where people are using Uber / Lyft to get a ride to and from the game. The surge pricing is designed to bring an increase in supply to an area that lacks adequate supply for the demand.

The latter statement makes the high demand and low supply periods equivalent for the purposes of surge pricing - in both cases there is a mismatch between demand and supply, with demand being higher than supply.

Considering this is a short term investor forum surging should be a really easy concept.... if demand outstrips supply the price goes up, uber is litterally a stock market for taxis. There's nobody at some central office going "oh it's raining, let's crank it up to 3x". It's all about how many passengers request a ride and how many active cars there are. Only difference is uber sets minimum prices. If TN can be lower than that then well, Uber's ****ed. I wish Uber would IPO so I could short it.

 

[Lapin7]

At the moment you can earn like €200 per day through a P2P Car rent service called SnappCar. Like that the Tesla S becomes more then affordable.

Tesla has of course seen this opportunity as well and they are creating an App that allows to add your car to their shared fleet of cars. If you want to make a ride, then you send a request to the nearest Tesla. The Tesla drives autonomously drives to your pick up point. You jump in and take your Ride. At arrival you jump out and pay.

You can imagine the enormous potential of this disruption in mobility services. And also the road blocks put into place to avoid this beautiful dream. Imagine for the violent reaction of establisment on for example:
- disruption of mobility services (public and private transport services for goods, persons, animals and things). Think of buying miles instead of vehicles.
- disruption of car insurance, because of metered fees based on car behavior and traffic circumstances
- disruption of oil industry
- disruption of car constructors
- disruption of electricity distribution
- finally a solution for CO2 reduction
- .......

Now the problem is that it's going to be a centralised service called Tesla Networks. It will be attacked for all kind of miss-used reasons:
- privacy concerns. One party, Tesla, holds all information about all mobility in the world.
- safety concerns. Laws that forbid autonomous driving cars
- media will report about every Tesla car accident
- financial monopoly of Tesla
- all kind of centralised restrictions by Tesla on usage of Tesla cars (maybe they decide that you can not go to certain places, freedom of movement will be constrained)
- cyber criminality with concern to connected autonomous cars
- ....

The solution is of course to create an unstoppable crypto service through a decentralised and distributed system inspired by Bitcoin, like Ethereum, Synereo, Byteball.

What kind of info can give Tesla about these restrictions, guarantees, dangers, possibilities, solutions

Thanks for reading,

 

[Bankroetlama]

The Tesla car sharing network/transport as a service isn't anything new or unique. Every car manufacturer + Google + Uber is racing to develop the first fully self driving car and have the monopoly. They won't sell a car one time for 3-4%, they will sell the service for 10-20% margin continuously

 

[Green Pete]

The Tesla car sharing network/transport as a service isn't anything new or unique. Every car manufacturer + Google + Uber is racing to develop the first fully self driving car and have the monopoly. They won't sell a car one time for 3-4%, they will sell the service for 10-20% margin continuously

Neither google nore Uber actually makes cars. I find it unlikely either one will ever sell cars. I would guess they are buying them and then retrofitting with hardware or heaving no them built to custom order and then putting them on the road prioritarily. This is hugely capital intensive.

Tesla can sell the cars, establish a network and make a profit on it while taking market share without spending any real additional capital, they just keep doing what they've been doing. Sell cars.

If uber wanted to replace its 300,000 drivers lets says that take with self driving cars (and the expensive looking system on top of Uber's cars now) let's say that's 70,000 a car. That's a 21 billion dollar capital investment for a company that doesn't actually have any real assets.

I think the tech industry (and those enthusiastic about it) has become so disconnected with hardware costs. Tesla is really the only tech company that has come out of Silicon Valley that has seen manufacturing as an integral part of their goals.

 

[Bankroetlama]

Big news
Volkswagen is launching on-demand shuttle services in 2017 to take on Uber

 

[EinSV]

Reviving this old thread to add a perspective from Gene Munster, who came out with a note recently on the Tesla ride sharing opportunity where he concluded:

• While a long ways away, we believe there’s a greater than 50% chance that Tesla will operate a ride-sharing fleet by 2023.
• The fleet could add $2B to $6B in high margin revenue for Tesla starting in 2023. Tesla does not make money today.
• Separately, we believe there is also a greater than 50% chance that Tesla will allow Tesla owners to add their vehicles to the fleet part-time. While only a small percentage of owners will opt in, the financial reward is measurable and could earn the owner enough to pay their vehicle lease.
• We believe Tesla could have between 4% and 10% of the U.S. ride-sharing market in 2023.
• If several key obstacles are overcome, most notably achieving autonomy, Tesla will have a handful of substantial advantages over competitors in the space.

Tesla’s Multi-Billion Dollar Ride-Sharing Opportunity | Loup Ventures

To model the business opportunity, Munster runs several scenarios but for his mid-range scenario he assumes Tesla manages about 100K 3+ year-old cars in the portion of the fleet it operates and generates $32K per car in income per year. He also assumes that 10% of owners participate in the network, and that Tesla receives 10% of the ride sharing income from the owner operated fleet. Assumptions Worksheet.pdf He also assumes that a total of about 4M Tesla self driving cars are on the road by then.

Using these numbers, the Tesla fleet would generate about $3.5 billion in income (not revenue) in 2023.

Munster tends to use conservative numbers and I view these estimates as conservative. For example, using the figures above Tesla could increase its ride sharing income to $6.7 billion per year by adding 100,000 newly manufactured cars into the portion of the fleet it manages at $32,000 per year in income. Also, since under his model owners could essentially pay for their lease by putting their cars into the Tesla network, a significant number of people who could not otherwise afford a Tesla could choose to get a "free" Tesla by putting it into the network.

The main issue is achieving autonomous driving at a level that can be used in a Tesla network in a significant number of regions.

 

[ValueAnalyst]

Reviving this old thread to add a perspective from Gene Munster, who came out with a note recently on the Tesla ride sharing opportunity where he concluded:

• While a long ways away, we believe there’s a greater than 50% chance that Tesla will operate a ride-sharing fleet by 2023.
• The fleet could add $2B to $6B in high margin revenue for Tesla starting in 2023. Tesla does not make money today.
• Separately, we believe there is also a greater than 50% chance that Tesla will allow Tesla owners to add their vehicles to the fleet part-time. While only a small percentage of owners will opt in, the financial reward is measurable and could earn the owner enough to pay their vehicle lease.
• We believe Tesla could have between 4% and 10% of the U.S. ride-sharing market in 2023.
• If several key obstacles are overcome, most notably achieving autonomy, Tesla will have a handful of substantial advantages over competitors in the space.

Tesla’s Multi-Billion Dollar Ride-Sharing Opportunity | Loup Ventures

To model the business opportunity, Munster runs several scenarios but for his mid-range scenario he assumes Tesla manages about 100K 3+ year-old cars in the portion of the fleet it operates and generates $32K per car in income per year. He also assumes that 10% of owners participate in the network, and that Tesla receives 10% of the ride sharing income from the owner operated fleet. Assumptions Worksheet.pdf He also assumes that a total of about 4M Tesla self driving cars are on the road by then.

Using these numbers, the Tesla fleet would generate about $3.5 billion in income (not revenue) in 2023.

Munster tends to use conservative numbers and I view these estimates as conservative. For example, using the figures above Tesla could increase its ride sharing income to $6.7 billion per year by adding 100,000 newly manufactured cars into the portion of the fleet it manages at $32,000 per year in income. Also, since under his model owners could essentially pay for their lease by putting their cars into the Tesla network, a significant number of people who could not otherwise afford a Tesla could choose to get a "free" Tesla by putting it into the network.

The main issue is achieving autonomous driving at a level that can be used in a Tesla network in a significant number of regions.

I second that "Munster tends to use conservative numbers and I view these estimates as conservative."

For 2023, I have 20% of owners participating with Tesla taking 25% revenue share at 50% gross margin = $13.5 billion gross profit.

After OpEx and Taxes, comes to about double Munster's estimate.

 

[Crowded Mind]

I second that "Munster tends to use conservative numbers and I view these estimates as conservative."

For 2023, I have 20% of owners participating with Tesla taking 25% revenue share at 50% gross margin = $13.5 billion gross profit.

After OpEx and Taxes, comes to about double Munster's estimate.

Why 50% gross margin?

 

[GoTslaGo]

Why 50% gross margin?

Possibly since it’s mainly software?

 

[ValueAnalyst]

Why 50% gross margin?

Conservative estimate for a software business. Tesla Network will essentially be an app where Tesla owners meet riders. Facebook, for instance, has a gross margin above 85%, which is in-line with other pure software companies. Note that I am not projecting any content-related revenue, so Netflix is not a good comp. Twitter could be a comp, and it has a gross margin close to 70%. What are your thoughts?

 

[Crowded Mind]

Possibly since it’s mainly software?

Conservative estimate for a software business. Tesla Network will essentially be an app where Tesla owners meet riders. Facebook, for instance, has a gross margin above 85%, which is in-line with other pure software companies. Note that I am not projecting any content-related revenue, so Netflix is not a good comp. Twitter could be a comp, and it has a gross margin close to 70%. What are your thoughts?

I don’t have an estimate, but 50% seemed low to me. Data costs could roll into COGS here, anything else that would add up to $13.5B in costs?

 

[ValueAnalyst]

I don’t have an estimate, but 50% seemed low to me. Data costs could roll into COGS here, anything else that would add up to $13.5B in costs?

Just being conservative, that’s all. I wouldn’t expect Tesla network to impact stock price until after 2020 anyway.

 

[Crowded Mind]

Just being conservative, that’s all. I wouldn’t expect Tesla network to impact stock price until after 2020 anyway.

Agreed, that’s why I don’t have an estimate. Yet. I think if I were erring towards conservatism, I’d do it on the revenue line rather than gross margin though, justified by assumed competition driving lower prices or market segmentation.