FSD Subscription

[camthehombre]

I read through most of the comments but not all so maybe this has been visited, but I think one thing to take into account would be how much it actually costs Tesla to produce the FSD chip and to install in. I have no idea what the numbers are, so I'm just going to make them up. Let's say $500 for the chip and $500 for labor time to install.

There would then be factoring in how long people would keep it for, and what percent of people would go for it, plus those who would be open to a subscription when the $7,000 option was never on the table. Maybe I'm looking at it the wrong way, but there is always a risk of loss. What people would really opt for HW3 just to get the hardware instead of the software side of things? Sure there will be some people, but the chance of that is probably low, and if they did, the next person that gets the car might opt in and the hardware is now there. But is there really much benefit to having HW3 without the software side of things? Minus someone who really wants to see stoplights, stop signs, road markers and garbage cans.

I think something like $100 a month could be reasonable. Maybe they will have a minimum time, but I just don't know if they would. At this cost, it would allow Tesla to break even after 10 months, and anything above this they are actually making more money than they would have if they didn't do the install. So even if 10% of people went to get the hardware upgrade and then stopped after one month, anyone else who opts in could offset that. On top of that, if people already have HW3 but weren't willing/able to fork out $7,000 could go for even just a month, they are already getting 100% of profits they weren't getting before.

All that being said, maybe that's just hopeful thinking from someone with HW 2.5 and EAP. None of us really know because no other auto manufacturer has anything like this as an option. We are all wildly speculating.

 

[nayr14]

$7k for the current state of FSD is way to much. For actual "full self-driving", I believe new sensor packages will need to be created that may or may not retrofit onto the current Model 3. Even with HW3, there are still phantom braking issues on the interstate... they have a long way to go.

The other issue with paying up front is insurance. You pay the extra to insure the value of FSD, but that value deprecates with the vehicle. If you total your 2yo Model 3 with FSD, you'll probably be looking at ~$3500 from your insurance, less in the future. This isn't sustainable and I don't want to pay insurance an on extra $7k worth of software.

Subscription makes a lot of sense (and Tesla probably sees this now as an insurer in CA). There are many ways they can pitch it, however it's complicated by those who have already paid $7k at purchase. I'm hoping they will break down FSD into two options - "Basic" and "Advanced". Basic will have standard summon and highway lane change at around $30/mo. Advanced would have capabilities including the city stoplight/stop sign suite they are working on now, as well as the future features they are promising with FSD.

More and more vehicles are being offered with items that are part of FSD. The Hyundai Sonata has the equivalent of basic summon (RSPA) and now Cadillac's Super Cruise has highway auto-lane changing abilities. If they want to stay on top, they need to push these items into AP or offer them at a low subscription price. It somehow needs to be restructured.

 

[ElectricIAC]

When would that license "die"? If it can be passed on forever it would be bad for Tesla.

Dies with the owner I suppose.

 

[MP3Mike]

Dies with the owner I suppose.

I could agree with that. Or the other option is that when the person it is assigned to dies it reverts to being tied to the car. At least that way their family would get to keep use of the license for as long as they have the car. But it couldn't be transferrable to another person.

 

[DrtyJrze]

I could agree with that. Or the other option is that when the person it is assigned to dies it reverts to being tied to the car. At least that way their family would get to keep use of the license for as long as they have the car. But it couldn't be transferrable to another person.

Another option could be total time of Tesla ownership. So say the license was good for 10 years. So an FSD license could be used on one car for 10 years or 5 cars for 2yrs each, etc. Basically you use up the license, but only while currently owning a Tesla.

That would encourage brand loyalty and new vehicle purchases.

 

[trm2]

If Tesla wants to keep selling FSD at 7k or more the monthly subscription cost will have to be pretty high. They could switch to a subscription model ONLY and have a lower monthly cost. The Tesla thing to do would be to announce that it will only be subscription going forward and the price will be $X per month where X is an “introductory price” that will go up. Then, try to make a bunch of people rush to purchase FSD while they can still get it as a “lifetime (of the car) price” because they will no longer be able to lock down the “cheap“ $7k price. Naturally, two months later, a full purchase will be available once again at who knows what price.

 

[DrtyJrze]

If Tesla wants to keep selling FSD at 7k or more the monthly subscription cost will have to be pretty high. They could switch to a subscription model ONLY and have a lower monthly cost. The Tesla thing to do would be to announce that it will only be subscription going forward and the price will be $X per month where X is an “introductory price” that will go up. Then, try to make a bunch of people rush to purchase FSD while they can still get it as a “lifetime (of the car) price” because they will no longer be able to lock down the “cheap“ $7k price. Naturally, two months later, a full purchase will be available once again at who knows what price.

Hahahaha, definitely sounds like some BS Tesla would do.

Only problem is that the subscription only model would still be compared to the established price and therefore would have to be high in order to make people “rush” to pay the $7k upfront.

Here’s an option: What if they did a subscription service with an “Upgrade” fee of $4k and monthly payments of $150 with a 2 year agreement?

You can play with the numbers a dozen different ways to meet the financial needs of subscribers but still make buying upfront the cheapest overall option.

 

[HyperionMark]

This may have been mentioned already...
What about setting the monthly price a little higher (maybe $200ish) but that money gets taken out of the total cost should you choose to purchase it outright. For example, you pay the subscription for 3 months and then decide to purchase the full package outright after that. Based on the current $7k, the outright price would then be down to $6400.

 

[willow_hiller]

As some people touched on what Tesla owners are willing to pay, someone put together a poll on Reddit:

POLL: What would you pay for a monthly FSD subscription? : teslamotors

Over half of respondents say $10-$49 per month is what they would pay. About 80% are $10-$49 and $50-$99 together.

 

[drtimhill]

I read through most of the comments but not all so maybe this has been visited, but I think one thing to take into account would be how much it actually costs Tesla to produce the FSD chip and to install in. I have no idea what the numbers are, so I'm just going to make them up. Let's say $500 for the chip and $500 for labor time to install.

There would then be factoring in how long people would keep it for, and what percent of people would go for it, plus those who would be open to a subscription when the $7,000 option was never on the table. Maybe I'm looking at it the wrong way, but there is always a risk of loss. What people would really opt for HW3 just to get the hardware instead of the software side of things? Sure there will be some people, but the chance of that is probably low, and if they did, the next person that gets the car might opt in and the hardware is now there. But is there really much benefit to having HW3 without the software side of things? Minus someone who really wants to see stoplights, stop signs, road markers and garbage cans.

It's not a chip replacement, its the entire computer, so it will be way more than $500.

 

[DrtyJrze]

As some people touched on what Tesla owners are willing to pay, someone put together a poll on Reddit:

POLL: What would you pay for a monthly FSD subscription? : teslamotors

Over half of respondents say $10-$49 per month is what they would pay. About 80% are $10-$49 and $50-$99 together.

The responses and reasoning are just as ridiculous there as they are here, if not worse. Doesn’t add much value to this thread.

 

[HyperionMark]

The responses and reasoning are just as ridiculous there as they are here, if not worse. Doesn’t add much value to this thread.

I agree. No way on earth it is less than $50 per month. Would be surprised even if less than $100.

 

[S4WRXTTCS]

The responses and reasoning are just as ridiculous there as they are here, if not worse. Doesn’t add much value to this thread.

I think what we're seeing is the difference between what we're willing to pay now versus what we're willing to pay when it does more.

Asking someone to pay more than $50 a month for what it currently does over basic AP is asking a lot. I don't think there would be a lot of takers if released today at over $100/month. Maybe short time "try it out" ones, but not a lot of people who stick with it.

We already know what Tesla charges for a monthly FSD subscription because that's essentially what a lease is. The cost for FSD is approximately $100 a month in 36 month lease. That makes sense because it's actually what a typical $7K option that has a 50% depreciation over those three years would cost.

It's not surprising to me that 80% of the responders to the poll aren't willing to pay more than what it is now.

The justification for paying $7K upfront was not what it can do today, but what it can do tomorrow. That's what justifies the $100 a month. It allows us to get locked into that price while FSD gets more expensive. The risk is it never really amounts to being worth $100 a month.

For me it gave me HW3, and likely at least 2 years of SW updates with HW3 before HW4 (or some life event) ruins my fun. Whether I made the right call remains to be seen. I don't think NoA, Smart Summons, or Autopark are worth anything. They're all very much in the "this sorta, kinda works sometimes" phase where they should be paying us to use them to report back what works, and what doesn't work.

 

[S4WRXTTCS]

I agree. No way on earth it is less than $50 per month. Would be surprised even if less than $100.

There are two questions to ask.

The first question to ask is what people are willing to pay. The poll shows you what they would be willing to pay so I don't see much point in questioning it.

The second question to ask is what the minimum price that Tesla can sell it for. I completely agree with both of you that it won't be less than $100 a month. We already know that's the minimum because that's what Tesla charges for FSD in a lease.

As a result there is at least a $50 gap between what people are willing to pay, and what Tesla can sell it for.

My expectation is that Tesla simply won't sell FSD as a subscription until the gap is eliminated. Or they'll just sell it at $100 month like what they currently do (if rolled into a lease), but heavily warn people that FSD subscription pricing will likely go way up in the future. So it's best to lock the pricing by buying it or adding it to the lease.

I also expect the subscription pricing to only apply to HW3 vehicles as they don't require any HW upgrade. The big payoff for a lot of us with FSD was the HW upgrade. To me that's always going to be a least $2K out of the $8K (EAP+FSD) that I paid.

 

[CertLive]

I think FSD subscription is a great idea. Personally I might want to use it while I was going on a long drive across Europe. How much per month. Well consider the fact you cant be held back by people who paid $7000, you took this at the time so it should have no value link on the result for subscriber options change is change. FSD is selling on the lower percentage of 40% my guess. I would imagine it is then much more of a value proposition to split the features up and add them to a monthly no contract subscription. I think for current owners and for new owners when you buy on the website you should be able to tick each option for a breakdown price as follows:

Monthly:
Premium Features £9.99
Self Parking - £4.99
FSD - £9.99

Potential per customer over a 3 year lease: £898.92
8 year warranty £2397.12

Now that off probably 85% of your customers or one of the above would be a lovely revenue stream to extend from the next car to the next car. Better much much better.... £5800 for FSD currently is way too high for us being testers. Subscription based gives each user flexibility on their options and enjoyment of the car.

 

[DrtyJrze]

I think FSD subscription is a great idea. Personally I might want to use it while I was going on a long drive across Europe. How much per month. Well consider the fact you cant be held back by people who paid $7000, you took this at the time so it should have no value link on the result for subscriber options change is change. FSD is selling on the lower percentage of 40% my guess. I would imagine it is then much more of a value proposition to split the features up and add them to a monthly no contract subscription. I think for current owners and for new owners when you buy on the website you should be able to tick each option for a breakdown price as follows:

Monthly:
Premium Features £9.99
Self Parking - £4.99
FSD - £9.99

Potential per customer over a 3 year lease: £898.92
8 year warranty £2397.12

Now that off probably 85% of your customers or one of the above would be a lovely revenue stream to extend from the next car to the next car. Better much much better.... £5800 for FSD currently is way too high for us being testers. Subscription based gives each user flexibility on their options and enjoyment of the car.

Thank you for adding to the convo and actually using numbers to support your thoughts.

Tesla isn’t being “held back” by people who paid $7k. Tesla is held back by the revenue they would lose.

If the uptake is 40% for FSD (your assumption) Tesla would be financially destroyed if they introduced a subscription service for $10, $15, $50, or $100 a month. As mentioned previously, every one person who pays the $7k equals dozens of people who “might” pay $15 a few months out of the year. So the FSD take rate would have to be about 4% before we would even began to have a discussion for prices that low. Is there anyone here who thinks only 4% of Tesla vehicles have FSD? If not, PLEASE stop with the $15/month talk. It’s ludicrous.

For the 5th time, this is a simple math equation. Can ANYONE provide a business case and not just end user pie in the sky fantasy prices? I’d love for Elon to thank us all for being beta testers and give us our cars for free, but that’s as likely to happen as a $10 subscription.

Oh, and electric, looking forward to seeing you disagree with logic and math again. It makes me feel all warm and fuzzy inside.

 

[drtimhill]

I agree. No way on earth it is less than $50 per month. Would be surprised even if less than $100.

First, let's assume Tesla do indeed decide on a subscription model (by no means certain). How do they choose a price-point?

-- They would somehow have to factor in the HW3 upgrade, which is not cheap. This means there is either an up-front "stop" cost (maybe $1500-$2000), or a minimum subscription period to cover that cost (and more).
-- Then there is the usual "maximize area under curve" model, where they want to set the price to maximize total dollar return. This is tricky; typically you would take into account well-heeled buyers, but in this case many of those have already purchased the FSD package. So your audience is reduced to people who are more limited in available $$, which points to a lower price point.
-- BUT, you then have to factor in a potential back-lash from those who already have paid $7000 (or similar) for FSD. Assuming you keep the car for 5 years, and maybe can add $2000 (which is high) to the resale value of the car because it has FSD, you are looking at a minimum monthly cost of $83.

Sure, you could ignore those people who have already purchased FSD, but dropping the price on something before you even ship it (completed FSD) is NOT going to please existing owners, and these are, by definition, premium customers for Tesla.

I don't see how Tesla could charge less than $80-$100 per month. Even then the math for them seems dubious to me.

 

[Knightshade]

I don't see how Tesla could charge less than $80-$100 per month. Even then the math for them seems dubious to me.

There's some cases where it's a long term subscription you're locked into, and don't own it at the end, where they could maybe do that.

$75 a month, for 5 years, and you own nothing at the end of it.

That's $4500 for renting him FSD. The guy who didn't want to spend 7k saves money, but Tesla gets 4500 they wouldn't otherwise, and they get to sell it to the next owner too without needing to take possession of the car back to do it.

Or maybe even offer that guy a $4000 "buy" option at the end if he thinks he's actually keeping the car a few more years- so he jumps on it because he can now recoupe that when he sells- but Tesla still got $1500 more from him than a dude who bought it for 7k upfront?


What I don't see, at nearly any price anyone would pay, is a month-to-month no commitment subscription- unless Tesla is totally abandoning the idea of getting past L2 driver aids.

 

[ElectricIAC]

Non-equity option would be okay but perhaps at a “residualized” rate like on a lease.

 

[CertLive]

The cars come with the hardware anyway. Subscription at a low price would bring in way more revenue as more of the cars are sold. You see unless Tesla is selling the cars at a loss in hope of more FSD sales to bridge the gap then better to have people connected that way than have it all ride along the road dormant. FSD is nowhere near complete yet and a lot of people including myself have a hard time putting money down on a feature that may take years and its quite possible my model 3 doesn't have the hardware for full autonomy (not discounting retro fit). Full self driving is worth $7000 but even subscriptions could rise when it is complete. For now make some money off the majority of the cars, let more people test it and at a low price to reward the testers.