Tesla Reservations Model and Demand Generation

[vgrinshpun]

Sorry for the brief tangent... but where and when and how was this test drive event publicized and held? I have been seeking out sources for Philly Tesla events (including the KoP store opening) but thus far my Google-fu has been extremely weak in this regard.

Is there a group somewhere that I just don't know the secret handshake for?

I don't know if this event was publicized. Last sunday I just happend to fill in a request for test drive on line, thinking that I would get test drive at the NJ Paramus store. I've got call back next day from Tesla employee from the Paramus store, indicating that they are planning test drive event in Philadelphia.

The good news is that according to the King of Prussia store manager with whom I spoke after the test drive, the King of Prussia store is going to open on May 10 - 12.

 

[luvb2b]

a quick comment on semantics. i don't think the reservations model i posted can change. that's because the model i put forth is an accounting fact in the sense that:

prior quarter ending reservations + this quarter's new reservations - this quarter's production - this quarter's cancellations = this quarter's ending reservations

after pondering for some time, i don't believe there is another valid reservations model that doesn't have this equation at its core.

therefore i believe the only differences in opinions can be in the specifics of what is used for new reservations, production, and cancellations.

- - - Updated - - -

Regarding the cancellations, the model does include them. Calculations are based on EM indicating 5 month wait on Feb 20 during the Q4 2012 call, so cancellations up to Feb 20 are accounted for. The wait time between the ordering/puting down deposit and finalizing was reduced to 2 weeks in the second part of March, so cancellation virtually went to zero. So the only period that is not covered is between Feb 20 to the mid-March. Once again cancellation rate was very low as wait time was drammatically decreased in the second half of Q1.

i don't believe this is correct. a number of reservations were sitting on the books as deferrals. some percentage of these cancel and some hang on. the 5 month average wait is a global figure, north american times were lower. the wait time you're describing is north america, but you've still got a huge percentage of global reservations on the books that will likely see elevated cancellation rates just as the north american reservations did at the start. understating the cancellation rate will throw everything way off.

 

[ModelS8794]

Create an account in Tesla's website. They sent me an email last week about it.

Thanks, I am an owner so have an account there. Maybe went to spam mail, or maybe current owners were not included? Seems unlikely, I dunno...

 

[vgrinshpun]

I don't believe this is correct. a number of reservations were sitting on the books as deferrals. some percentage of these cancel and some hang on. the 5 month average wait is a global figure, north american times were lower. the wait time you're describing is north america, but you've still got a huge percentage of global reservations on the books that will likely see elevated cancellation rates just as the north american reservations did at the start. understating the cancellation rate will throw everything way off.

Have to disagree with you. The 5 month wait that EM mentioned during the Q4 2012 CC definetely did not account for the EU reservations as he was speaking about situation in US. NA times were not lower, EM was adressing exactly this question in the CC. According to his explanation the shorter wait time for performance models was not indicative of the total backlog, as one would need to consider wait time for all configurations of MS. Take another listen to the call - as far as I remember this was addressed in the first question to EM.

Note that my calculation is done for US reservations only, and European cancellation rate is irrelevant. As per my previous explanation US cancellations for the period when they were substantial are accounted for.

As a side note, during his introduction remarks at the Geneva Auto Show EM indicated that Tesla plan to deliver "5,000 - 7,000 cars, may be slightly more" to Europe in 2013. He also indicated that if somebody in Europe wants to get the car in 2013 they need to place reservation "fairly soon". During the Q4 2012 call EM (or GB) indicated that out of 15,000 net reservations on the books about 25% are outside Northe America

 

[luvb2b]

Have to disagree with you. The 5 month wait that EM mentioned during the Q4 2012 CC definetely did not account for the EU reservations as he was speaking about situation in US.

hmm, maybe i am wrong i still think there's something wrong with your conclusions. i'm comfortable with what i have so didn't really dig into the details to figure out where we differ. for now we can agree to disagree.

 

[vgrinshpun]

@vgrinshpun

What I don't like about your calculations is the way you arrive at the reservations rate. EM has given a ballpark picture of backlog in late Feb, but the reservation rate in your model depends critically on exactly what that figure is. Subtract 5% from that figure, and the reservation rate that comes out of the model drops from 35 to 23 cars per day. And I can assure you that such ballpark figures have much higher uncertainty than +/- 5%!

In order to have an alternate reservation rate information, I've gone back to the Reservation Tally threads here and over the Tesla Motors Forum. Based on information included at these threads average rate for the Jan 1 through March 3 was 30 cars per day. This reservation rate yields production rate of 584 cars/week vs. 617 cars/week for the previously used reservation rate of 35 cars/day.

The bottom line is that Tesla most likely increased their production in Q2 to significantly more than 500 cars per week. Note that both results are conservative as the calculation is using flat 30 or 35 cars/week reservation rate, while we know that reservation rate was significantly higher during the warmer weeks of April.

The major takeaway is that this increased production is not a known fact yet, and when it is disclosed during the Q1 CC there is potential for a very strong market reaction.

- - - Updated - - -

hmm, maybe i am wrong i still think there's something wrong with your conclusions. i'm comfortable with what i have so didn't really dig into the details to figure out where we differ. for now we can agree to disagree.

We actually do not have competing models at all.

You've gone to a great lengths to produce reasonably accurate information on the backlog of reservations in the absence of comprehensive information required to precisely calculate this backlog.

I am not trying to derive an alternate way to determine the backlog of reservations at all. I am just grabbing a number to show that once information of the backlog of reservations and projected wait time are known, one can calculate the corresponding production rate (to go through known amount of reservation within set period of time).

The surprising result of this exercise is that any reasonable projection on backlog of reservations, when combined with the delivery times shown on Tesla Motors Design page invariably point to the fact that Tesla factory is currently operating at about 600 or more cars per week. Tesla disclosing this along with the corresponding increased projections for Q2 and 2013 production (upwards to 29,500 cars) could be the match that will start wild fire after the Q1 CC.

 

[DonPedro]

Ultra-fast EU finalization

It is quite striking how quickly Tesla are pushing out finalization buttons in Europe these days. In Norway, all reservations up to at least P2681 have got the button, whereas in the rest of EU we have heard about the button at P23XX. This means that button has gone out to approx:

Sigs: 500?
Rs: 400?
Ps: 2500?

Subtract some prior cancellations and sig upgrades, and we may be talking around 3000 reservation holders getting the button. The latest have been told to expect the car in Sep/Oct.

The rush has been apparent for those finalizing. According to reports, paperwork takes many days to return, and staff admit things are chaotic.

What could be reasons for pushing the finalizations so fast? Some ideas:

• Plans to announce a solid amount of "firm orders" at Q1 presentation
• Attempts at streamlining production by grouping similar orders
• Greater-than-expected cancellation rates
• Greater-than-revealed production rates
• "We simply want to know where we stand"

Anyone care to speculate?

 

[Citizen-T]

The more cars they can put on a boat, the cheaper it is to ship each car.

 

[DonPedro]

The more cars they can put on a boat, the cheaper it is to ship each car.

All indications are that the cars will be shipped in 4-car containers, in which case there are no significant economies of scale to the sizes of individual shipments.

 

[vgrinshpun]

It is quite striking how quickly Tesla are pushing out finalization buttons in Europe these days. In Norway, all reservations up to at least P2681 have got the button, whereas in the rest of EU we have heard about the button at P23XX. This means that button has gone out to approx:

Sigs: 500?
Rs: 400?
Ps: 2500?

Subtract some prior cancellations and sig upgrades, and we may be talking around 3000 reservation holders getting the button. The latest have been told to expect the car in Sep/Oct.

The rush has been apparent for those finalizing. According to reports, paperwork takes many days to return, and staff admit things are chaotic.

What could be reasons for pushing the finalizations so fast? Some ideas:

• Plans to announce a solid amount of "firm orders" at Q1 presentation
• Attempts at streamlining production by grouping similar orders
• Greater-than-expected cancellation rates
• Greater-than-revealed production rates
• "We simply want to know where we stand"

Anyone care to speculate?

Very interesting data - let’s try to work through the numbers to see if it is consistent with Tesla’s increased production rate that I’ve hypothesized couple of posts above.

During the Geneva Motor Show remarks EM indicated that Tesla will deliver 5,000 to 7,000 or a “bit more” cars to Europe in 2013. Assuming that back when EM was making these remarks the plan was to produce 500 cars/week in July and August and 650 cars/week starting in September, the total production goal was 8 x 500 +18 x 650 = 15,700 cars (Q3 and Q4). Therefore the intent was to send 32% (5,000 / 15,700) to 45% (7,000 / 15,700) of all produced cars to Europe. The average of two numbers works out to approximately 38%.

Based on information in Don Pedro’s post above there is total of about 3000 reservations with the latest expected delivery of 3 months after the start of European deliveries in July. If one assumes that these 3,000 cars to be delivered to Europe correspond to a 38% of cars produced during the same period of time, the total production would be 3000 / 0.38 = 7895 cars, or 7895 / 13 = 607 cars a week.

So this data is consistent with the hypothesis that Tesla already increased or planning to increase production sooner than previously disclosed. By implication this means that Tesla sees improving margins and reservation rates.

The next week Q1 CC should be very interesting indeed.

 

[Benz]

The next week Q1 CC should be very interesting indeed.

It will be very interesting indeed.

 

[luvb2b]

I am not trying to derive an alternate way to determine the backlog of reservations at all. I am just grabbing a number to show that once information of the backlog of reservations and projected wait time are known, one can calculate the corresponding production rate (to go through known amount of reservation within set period of time).

i think i understand what you are saying. the problem is the reservations include deferrals which won't go into the production queue until they firm their orders. so taking your approach would overstate the production rate.

also it seems like you are completely ignoring an elevated cancellation rate, which will make your production figures way too high.

 

[vgrinshpun]

i think i understand what you are saying. the problem is the reservations include deferrals which won't go into the production queue until they firm their orders. so taking your approach would overstate the production rate.

also it seems like you are completely ignoring an elevated cancellation rate, which will make your production figures way too high.

I disagree with you saying that this approach ignores cancellations. As I explained before, the numbers I am using to estimate backlog prior to Feb 20 are net, so they account for cancellations. The cancellations between Feb 20 and mid-March had minimum cancellations as waiting period shrank, and there should be virtually no cancellation from mid-March, as the order is finalized within two weeks. Note that this calculation was done for NA only, without counting any European reservations.

Ultimately, however, we are in agreement that the reservation numbers that I use are estimates, as no precise information is available. That is why I call these musings a hypothesis.

I think that we both agree that Q1 looks good. If my hypothesis turns out to be correct, the Q1 CC will be stellar. There are only few days left until we know.

 

[luvb2b]

elon mentioned twice during the call that they are seeing an uptick in demand after the financing announcement. imo that should mean bullish commentary for the next quarter.

 

[DonPedro]

elon mentioned twice during the call that they are seeing an uptick in demand after the financing announcement. imo that should mean bullish commentary for the next quarter.

Good catch. I really respect what they did today - the admitting they made mistakes, the fast turnaround on the new version, and (not least) the constant keeping Tesla in the media. And, of course, I think they've launched a great financing product. I love Elon's vision: Going from top 1% to top 10% in terms of who would get the car.

 

[CapitalistOppressor]

@vgrinshpun

I think your logic is sound, and I think it would be just like Tesla to increase production capacity as EU deliveries draw near, to not lose momentum in deliveries in the US unless absolutely unavoidable. Your model also shows why it makes sense to assume that TMC could close all stores and still deliver cars at full throttle for the rest of the year.

What I don't like about your calculations is the way you arrive at the reservations rate. EM has given a ballpark picture of backlog in late Feb, but the reservation rate in your model depends critically on exactly what that figure is. Subtract 5% from that figure, and the reservation rate that comes out of the model drops from 35 to 23 cars per day. And I can assure you that such ballpark figures have much higher uncertainty than +/- 5%!

European demand is easy to calculate, since we still have the old reservation number system. It has been a quite steady rate of 11-12 cars per day (7 days a week). What will happen to that reservation rate when the Model S starts rolling in the streets here is anyone's guess, but it is certainly going to move upward! This increase will be supported by the increasing number of stores.

US demand is the #1 question right now. 2 per store per day seems like a good approach to understanding the potential. It is also very easy to use. The downside of that metric is that it does not say anything about how Tesla is doing in terms of capitalizing on that potential. As smorgasbord pointed out above, waiting times are a poor indicator of reservations, as TM probably have a lot of flexibility now to increase production temporarily to meet demand peaks.

I think the best thing to do would be for the TSLA investors at this forum to put some $ in a pool to hire someone to sit at the factory gates and count the cars rolling out. Knowing delivery rates would help A LOT wrt. knowing reservation rates. (I am only half joking here).

A couple of weeks ago I started thinking about what it would take to find a nearby property owner who would let us set up a web cam point at NUMMI, lol.

- - - Updated - - -

a quick comment on semantics. i don't think the reservations model i posted can change. that's because the model i put forth is an accounting fact in the sense that:

prior quarter ending reservations + this quarter's new reservations - this quarter's production - this quarter's cancellations = this quarter's ending reservations

after pondering for some time, i don't believe there is another valid reservations model that doesn't have this equation at its core.

therefore i believe the only differences in opinions can be in the specifics of what is used for new reservations, production, and cancellations.

- - - Updated - - -



i don't believe this is correct. a number of reservations were sitting on the books as deferrals. some percentage of these cancel and some hang on. the 5 month average wait is a global figure, north american times were lower. the wait time you're describing is north america, but you've still got a huge percentage of global reservations on the books that will likely see elevated cancellation rates just as the north american reservations did at the start. understating the cancellation rate will throw everything way off.

Very much, all of this. That is all.

- - - Updated - - -

Have to disagree with you. The 5 month wait that EM mentioned during the Q4 2012 CC definetely did not account for the EU reservations as he was speaking about situation in US. NA times were not lower, EM was adressing exactly this question in the CC. According to his explanation the shorter wait time for performance models was not indicative of the total backlog, as one would need to consider wait time for all configurations of MS. Take another listen to the call - as far as I remember this was addressed in the first question to EM.

Note that my calculation is done for US reservations only, and European cancellation rate is irrelevant. As per my previous explanation US cancellations for the period when they were substantial are accounted for.

As a side note, during his introduction remarks at the Geneva Auto Show EM indicated that Tesla plan to deliver "5,000 - 7,000 cars, may be slightly more" to Europe in 2013. He also indicated that if somebody in Europe wants to get the car in 2013 they need to place reservation "fairly soon". During the Q4 2012 call EM (or GB) indicated that out of 15,000 net reservations on the books about 25% are outside Northe America

As with Luvb2b, I don't have the time to really drill down into your model. But I have problems with the bolded comment. I'm pretty sure that the reservation numbers Tesla announced in February were as of Dec 31st and emphatically did NOT account for the large cancellation and deferral numbers that were shaken out of the system in January.

I empathize with your desire to play Kremlinology games with Elon Musk comments, because I've often done the same. But its my experience we need to back out and build off of actual data we know, and often a credible model will show that Musk is being "accurate", but in a way you wouldn't think of if you were using different assumptions when you interpret what he says.

- - - Updated - - -

It is quite striking how quickly Tesla are pushing out finalization buttons in Europe these days. In Norway, all reservations up to at least P2681 have got the button, whereas in the rest of EU we have heard about the button at P23XX. This means that button has gone out to approx:

Sigs: 500?
Rs: 400?
Ps: 2500?

Subtract some prior cancellations and sig upgrades, and we may be talking around 3000 reservation holders getting the button. The latest have been told to expect the car in Sep/Oct.

The rush has been apparent for those finalizing. According to reports, paperwork takes many days to return, and staff admit things are chaotic.

What could be reasons for pushing the finalizations so fast? Some ideas:

• Plans to announce a solid amount of "firm orders" at Q1 presentation
• Attempts at streamlining production by grouping similar orders
• Greater-than-expected cancellation rates
• Greater-than-revealed production rates
• "We simply want to know where we stand"

Anyone care to speculate?

Same thing they did in January. They are shaking the bush to see which reservations are real sales. Most reservation holders in the U.S. got the finalize button in January, with many still receiving their cars. It has nothing to do with production rates or streamlining, but is intended to reveal cancellations and deferrals, and they are doing it now because they will not have to report the results in the Q1 conference call.

 

[DonPedro]

A couple of weeks ago I started thinking about what it would take to find a nearby property owner who would let us set up a web cam point at NUMMI, lol.

Sign me up as footing part of the bill.

Same thing they did in January. They are shaking the bush to see which reservations are real sales. Most reservation holders in the U.S. got the finalize button in January, with many still receiving their cars. It has nothing to do with production rates or streamlining, but is intended to reveal cancellations and deferrals, and they are doing it now because they will not have to report the results in the Q1 conference call.

I agree. Thanks for your insight.

 

[CapitalistOppressor]

Very interesting data - let’s try to work through the numbers to see if it is consistent with Tesla’s increased production rate that I’ve hypothesized couple of posts above.

During the Geneva Motor Show remarks EM indicated that Tesla will deliver 5,000 to 7,000 or a “bit more” cars to Europe in 2013. Assuming that back when EM was making these remarks the plan was to produce 500 cars/week in July and August and 650 cars/week starting in September, the total production goal was 8 x 500 +18 x 650 = 15,700 cars (Q3 and Q4). Therefore the intent was to send 32% (5,000 / 15,700) to 45% (7,000 / 15,700) of all produced cars to Europe. The average of two numbers works out to approximately 38%.

Based on information in Don Pedro’s post above there is total of about 3000 reservations with the latest expected delivery of 3 months after the start of European deliveries in July. If one assumes that these 3,000 cars to be delivered to Europe correspond to a 38% of cars produced during the same period of time, the total production would be 3000 / 0.38 = 7895 cars, or 7895 / 13 = 607 cars a week.

So this data is consistent with the hypothesis that Tesla already increased or planning to increase production sooner than previously disclosed. By implication this means that Tesla sees improving margins and reservation rates.

The next week Q1 CC should be very interesting indeed.

My current thesis is that they are very close to their planned margins. Reservation rates are decent.

But you need to ratchet back the enthusiasm a bit. Delivery data as of a week and a half ago was looking a bit soft to support a 500/week production rate. I didn't do a full analysis because there isn't enough data, but it was probably somewhere between 400-500 in April. For a variety of reasons, but mainly because of the switchover to Multi-Red, it is hard to model the delivery flows for the few weeks bracketing April 1. They didn't finish backfilling until just recently, and there is a serious lag in the VIN data, both from people posting their deliveries and for my ability to analyze it.

The point is that there is a clear possibility that the announcement will be of current production in the 400/week range, and in a recent interview Elon said Tesla was producing 20k vehicles per year (which is 400/week). That's not the most likely possibility, but it is within the range of the actual data we have.

 

[Dr.Ling]

All I want to say is: Don't underestimate the sales potential in tiny Norway.

Practically everyone I know of talks about Tesla, and wants one too, and not because of me preaching! The Model S is the bargain of the century over here, and sales of electric cars are soaring. Even without the insanely tempting low-interest financing offered in the US the orders will be pouring in when people can see and feel the Model S for themselves.

 

[luvb2b]

All I want to say is: Don't underestimate the sales potential in tiny Norway.

Practically everyone I know of talks about Tesla, and wants one too, and not because of me preaching! The Model S is the bargain of the century over here, and sales of electric cars are soaring. Even without the insanely tempting low-interest financing offered in the US the orders will be pouring in when people can see and feel the Model S for themselves.

my norwegian is rusty but according to this tesla almost sold out norway for 2013 already.

http://www.tv2.no/underholdning/broom/tesla-model-s-utsolgt-i-2013-aatte-maaneders-ventetid-4039044.html?utm_source=Broom+Nyhetsbrevliste&utm_campaign=a868c81939-Broom_Nyhetsbrev_4_mai_20135_4_2013&utm_medium=email&utm_term=0_4a512f921e-a868c81939-225950125