General Discussion: 2018 Investor Roundtable

[Intl Professor]

Today we pick up our M3 at 1:00 pm. at Sacramento-Rocklin. So far in the delivery process the only glitches have been associated with corrections to our account on the web site. First, misspelling of wife's middle name on registration section. About a wait of a week, despite phone messages and e-mail before a delivery specialist called and fixed it on the fly—with warning it might take awhile for the website to show correction. It showed correction immediately. No wait. Then I got a call from a nice lady noting my driver's license had expired (old one posted in July) and politely asked me to photo copy of license which had been renewed in August. Did so and that day the web site changed—within hours.

Later I received a call from the local office asking if I wanted the car on Friday last instead of today but stuck with Sunday pickup as another party already free to drive with me or for me (my wife is afraid I would fiddle with the screen while driving). No problem. When I asked about parking she said, at first, we have a spot reserved for you and things should be less hectic on a Sunday. Yesterday I got an email reminder, generic, from the local store suggesting procedures in detail, register at first, then wait for assistance, then a five minute procedure, and a warning about parking some 100 yds distance at a local strip mall.

My only worries: peeing, there's only one public bathroom (although there may be porta-poties), parking, and I want a thorough inspection by yours truly and my driver and wife, which may take more time. Will post videos of crowd later if we can take them. Mind you, this center services over a million inhabitants within about 50 miles, maybe 1.5m. And Bonnie is close and, I guess, influential, so service may be unrepresentative.

At first contact I am always polite, note that as an investor I am happy they are busy but then that TMCers know you are over overworked and that more than most I will try to overcome my innate consumer interest and be more than usually patient. (Last, probably a lie. But I get good service.)

Turns out perhaps only five of us were receiving Teslas at that time. My alternate driver was late. At 1 p.m. sharp received a call from Tim asking for an update. Said we would have no problems with parking and paperwork would be ready. When we arrived he was waiting before a vacant parking spot near the entrance, took us in, checked with the guy at the door, and then escorted us immediately to a small room where paperwork was ready. After they copied, again as they were already posted, our driver's licenses, we were good to go. (20 m. for this?) Then handed off to Ryan who took us to a beautiful car and spent two hours showing how to do various things, making sure we had authenticated phones, profiles established, etc. Took about two hours. (I was very nervous, hadn't slept well, or it might have gone more swiftly. Ryan was nonetheless very patient, friendly, competent.)

Because I have limited movement of neck in leftward glances I damn near hit a car a year ago when I switched lanes. So my dear wife insisted we bring a shotgun to ride with me and I felt so dodgy I really wanted him to drive the two of us home while my wife drove separately with the Lexus LS430. The exterior mirrors are so much better than the Lexus I drove home anyway. Of course I have trouble with adjusting speed using regenerative braking and discovered I am used to driving with speed indicated in forward dash at wheel. TACC will take care of that. I was afraid to use it. I did notice something about acceleration. When I wanted to shift lanes, if I see a spot ahead, with an eye on a car I need to pass, you just point this rocket speed fantastically, and no problem. Crazy movement but actually so much safer in the M3 for an old fart.

Nearer home on city streets we were at a stop light next to a guy on the left with a really noisy motorbike. He took off like a shot and I punched it. I think we had more acceleration. This is by far the fastest moving rocket I've ever driven. The 2001 Lexus was fine in its day but you really have to crank up the engine. Nothing like the M3. You guys who want the performance version better be careful.

The Lexus does ride better and is quieter re road noise.

Very impressed with the service. There were more in the store when we left. I'm sure we lucked out on the Sunday delivery. I'll post more details about driving at over 80—years that is—on another thread.

Thanks for the support.

 

[adiggs]

I realize it's a pretty substantial database problem; at this point if I were there I would be arranging a Postgres/Citus migration. I don't know what they're actually using but it isn't up to the task. Admittedly Postgres/Citus wasn't an option two years ago.

As you, I too am an outsider looking in. Your assessment of a database problem is consistent with my own view of Tesla's internal information systems that I've been forming since my first interaction with the company in Spring '13, when I tried to persuade them that my new Roadster was really mine, and not the previous owners'. Service had to change the owner contact info every time we had the Roadster in for the first 2 or 3 visits (because the service DB kept being updated by the sales/marketing db, and service couldn't push an ownership change the other direction).

Anyway - this doesn't look like a technology problem to me. The way I read it, it doesn't matter whether they're using Postgres, SQL Server, Oracle, MySQL, or something else - it's a data structure / design problem and changing technologies isn't going to change the structure / design problem.

Which also means that this isn't going to be a quick fix, and greater efforts at quick fixes are more likely to result in bigger problems than fixes.

Though this is my field (database administration, data architecture, data integration, data management, etc..), I'm also an outsider; I don't have any direct knowledge. Only that I've had to solve similar problems where I did have direct knowledge and accountability for the system, and I'm projecting from that experience to what I've been seeing from Tesla for several years.

 

[aubreymcfato]

As you, I too am an outsider looking in. Your assessment of a database problem is consistent with my own view of Tesla's internal information systems that I've been forming since my first interaction with the company in Spring '13, when I tried to persuade them that my new Roadster was really mine, and not the previous owners'. Service had to change the owner contact info every time we had the Roadster in for the first 2 or 3 visits (because the service DB kept being updated by the sales/marketing db, and service couldn't push an ownership change the other direction).

Anyway - this doesn't look like a technology problem to me. The way I read it, it doesn't matter whether they're using Postgres, SQL Server, Oracle, MySQL, or something else - it's a data structure / design problem and changing technologies isn't going to change the structure / design problem.

Which also means that this isn't going to be a quick fix, and greater efforts at quick fixes are more likely to result in bigger problems than fixes.

Though this is my field (database administration, data architecture, data integration, data management, etc..), I'm also an outsider; I don't have any direct knowledge. Only that I've had to solve similar problems where I did have direct knowledge and accountability for the system, and I'm projecting from that experience to what I've been seeing from Tesla for several years.

This is why a collective doc written by you, @Troy, @neroden about the "probable" database issue could be helpful.
It will either picked up by Fred from Electrek or @ZachShahan from CleanTechnica, and if not you could post it on Reddit.
I'm sure we could make sure that Elon or some Tesla people will see it.

 

[avoigt]

At the end of the day regardless if we talk EVs or Battery projects the efficiency will make the difference and create an ROI that is hard to beat. Right now people still believe to go sustainable is more costly and not a viable option for most cases. That will change soon and may just have changed for utility battery projects like in Australia.

At that moment in time people will buy into sustainable energy because its more cheap for them. We will experience a huge new unexpected increase in demand once the public does realizes this.

The Australian project has proven already now that they are about to achieve a very quick ROI. That proof of concept will convince many others to invest into it because its more reliable, faster, batter ROI and finally good for the carbon footprint and the public image.

Tesla’s massive Powerpack battery in Australia cost $66 million and already made up to ~$17 million

 

[winfield100]

At the end of the day regardless if we talk EVs or Battery projects the efficiency will make the difference and create an ROI that is hard to beat.
At that moment in time people will buy into sustainable energy because its more cheap for them.
The Australian project has proven already now that they are about to achieve a very quick ROI. That proof of concept will convince many others to invest into it because its more reliable, faster, batter ROI and finally good for the carbon footprint and the public image.
Tesla’s massive Powerpack battery in Australia cost $66 million and already made up to ~$17 million

If you subscribe to RenewEconomy (daily emails) from Australia, you can follow it closer.
RenewEconomy

 

[neroden]

Did you try to file a report on your Tesla account and check escalate to exec? What about just telling someone in SC to escalate?

Twice each time for the USB problem, and untold numbers of times regarding the communications failures. Since none of that worked, I'm not going to waste my time doing it again for a failure I only know about secondhand (the delivery system data failures).

 

[neroden]

As you, I too am an outsider looking in. Your assessment of a database problem is consistent with my own view of Tesla's internal information systems that I've been forming since my first interaction with the company in Spring '13, when I tried to persuade them that my new Roadster was really mine, and not the previous owners'. Service had to change the owner contact info every time we had the Roadster in for the first 2 or 3 visits (because the service DB kept being updated by the sales/marketing db, and service couldn't push an ownership change the other direction).

Anyway - this doesn't look like a technology problem to me. The way I read it, it doesn't matter whether they're using Postgres, SQL Server, Oracle, MySQL, or something else - it's a data structure / design problem and changing technologies isn't going to change the structure / design problem.

Which also means that this isn't going to be a quick fix, and greater efforts at quick fixes are more likely to result in bigger problems than fixes.

Though this is my field (database administration, data architecture, data integration, data management, etc..), I'm also an outsider; I don't have any direct knowledge. Only that I've had to solve similar problems where I did have direct knowledge and accountability for the system, and I'm projecting from that experience to what I've been seeing from Tesla for several years.

Arguably the deeper problem has nothing to do with the computer system. It should have been possible for the service department to easily reach the sales/marketing people and to have them change the sales/marketing database data for your Roadster. It should have been, y'know, a half-day's work, sending in a copy of the paper work and getting it updated. *But it didn't happen*.

The underlying problem is that Tesla's internal communications is totally, utterly broken. Nonfunctional. Database communications replicate business logic based on face-to-face / telephone communications, and Tesla doesn't have the face-to-face / telephone communications working. Tesla's internal business logic is broken.

This would have been a quick fix if they'd done it right to start with, had managers and developed a suitable corporate culutre, but Musk arrogantly ignored the importance of management. They will have to spend billions trying to fix it and they may never fix it.

I still think the company's worth more than $300/share, but this is putting a cap on my valuation. Assuming they never fix this, it's going to destroy the brand, which won't matter until other companies produce in volume, but eventually they will -- people will take BAIC imports from China to avoid dealing with the Tesla communications nightmare. If they don't fix this, then at $600, I'm out.

 

[winfield100]

I still think the company's worth more than $300/share, but this is putting a cap on my valuation. At $600, I'm out.

ya gotta GTC for that? if it momentarily spikes? like IOM did in May 1996? $20-$100+-$20
or just reduce basis of remaining shares?

 

[AWDtsla]

Arguably the deeper problem has nothing to do with the computer system. It should have been possible for the service department to easily reach the sales/marketing people and to have them change the sales/marketing database data for your Roadster. It should have been, y'know, a half-day's work, sending in a copy of the paper work and getting it updated. *But it didn't happen*.

The underlying problem is that Tesla's internal communications is totally, utterly broken. Nonfunctional. Database communications replicate business logic based on face-to-face / telephone communications, and Tesla doesn't have the face-to-face / telephone communications working. Tesla's internal business logic is broken.

This would have been a quick fix if they'd done it right to start with, had managers and developed a suitable corporate culutre, but Musk arrogantly ignored the importance of management. They will have to spend billions trying to fix it and they may never fix it.

I still think the company's worth more than $300/share, but this is putting a cap on my valuation. Assuming they never fix this, it's going to destroy the brand, which won't matter until other companies produce in volume, but eventually they will -- people will take BAIC imports from China to avoid dealing with the Tesla communications nightmare. If they don't fix this, then at $600, I'm out.

These are problems that could be solved by 1 competent developer in a day. Don't know what the hold up is. I think at this point the employees have no idea and just quote large times to make people go away. Case in point, informed it would take up to 45 days to process unlimited supercharging referral. Just flip the ****ing bit man...

 

[neroden]

These are problems that could be solved by 1 competent developer in a day.

Yep, I could solve most of the problems personally, given access.

Let's be clear about this: Tesla is incompetent at computer stuff, and they're incompetent at communications. They've been incompetent for at least 5 years, and they've only gotten worse. As investors, we have to recognize that this is a severe weakness of the company, and that it will kill them as soon as inferior electric cars are produced in sufficient volume. (Which still isn't for at least two years.)

 

[Krugerrand]

Arguably the deeper problem has nothing to do with the computer system. It should have been possible for the service department to easily reach the sales/marketing people and to have them change the sales/marketing database data for your Roadster. It should have been, y'know, a half-day's work, sending in a copy of the paper work and getting it updated. *But it didn't happen*.

The underlying problem is that Tesla's internal communications is totally, utterly broken. Nonfunctional. Database communications replicate business logic based on face-to-face / telephone communications, and Tesla doesn't have the face-to-face / telephone communications working. Tesla's internal business logic is broken.

This would have been a quick fix if they'd done it right to start with, had managers and developed a suitable corporate culutre, but Musk arrogantly ignored the importance of management. They will have to spend billions trying to fix it and they may never fix it.

I still think the company's worth more than $300/share, but this is putting a cap on my valuation. Assuming they never fix this, it's going to destroy the brand, which won't matter until other companies produce in volume, but eventually they will -- people will take BAIC imports from China to avoid dealing with the Tesla communications nightmare. If they don't fix this, then at $600, I'm out.

Ok, it’s time you dropped this bone. There are hundreds of companies, nay thousands that have equally bad internal broken communications and it doesn’t put them out of business. It just makes them a bit more difficult to navigate for the consumer (an employee) if you have a problem you need addressed or if you want to talk to someone specific. But anyone with a little intelligence, patience, tack and perseverance can get what they need.

Let’s just start with every company that uses foreign call centers. None of them can solve sugar for you, push system changes, or for God sakes even connect you to a born English speaking person let alone a person with any power to resolve your issue.

The ability of people to effectively communicate with each other has been a historical issue; as in since the beginning of mankind. It’s not unique to Tesla and it won’t put them out of business anymore than TMC has been put out of business because we can’t get management to make changes here we’d like to see.

 

[Krugerrand]

Yep, I could solve most of the problems personally, given access

No you can not. You might be able to create a cohesive computer system but you won’t ever be able to solve the human element of communication, follow through, attention to detail et al... that is still required.

 

[neroden]

No you can not. You might be able to create a cohesive computer system but you won’t ever be able to solve the human element of communication, follow through, attention to detail et al... that is still required.

I just meant the technical support for the communication. The rest is a matter of creating a culture that cares about the internal communications, and getting rid of people who aren't on board with that. From five years of leaks, rumors, and comments, it seems like hardware engineering is actually serious about the communication feedback loop (good), software engineering isn't (bad), and the central call center doesn't even realize it's an issue (terrible).

 

[neroden]

Ok, it’s time you dropped this bone. There are hundreds of companies, nay thousands that have equally bad internal broken communications and it doesn’t put them out of business.

Actually, it often does put them out of business. Depends on whether they have a competent competitor. (Tesla doesn't yet.)

It just makes them a bit more difficult to navigate for the consumer (an employee) if you have a problem you need addressed or if you want to talk to someone specific. But anyone with a little intelligence, patience, tack and perseverance can get what they need.

I'll be pretty f***ing blunt here. When a customer is given the choice between the company where you have to spend an hour on the phone being kicked between different people none of whom know what they're doing just to get something you have a contractual right to, before you can get the simplest thing done...

...or a company where you call a secretary who knows what you're talking about and gets things scheduled for you immediately...

Which company are customers going to prefer?'

I realize that at the moment there basically isn't any other decent electric car, so, like Seinfeld's Soup Nazi, people will keep lining up for Teslas. But geeeeez, this is not a good reputation to have.

And for those who say "they'll fix it", they've had over five years and it's only gotten worse.

 

[Krugerrand]

I just meant the technical support for the communication. The rest is a matter of creating a culture that cares about the internal communications, and getting rid of people who aren't on board with that. From five years of leaks, rumors, and comments, it seems like hardware engineering is actually serious about the communication feedback loop (good), software engineering isn't (bad), and the central call center doesn't even realize it's an issue (terrible).

Your view as an outsider is clearly an incomplete view.

 

[mongo]

I'll be pretty f***ing blunt here. When a customer is given the choice between the company where you have to spend an hour on the phone being kicked between different people none of whom know what they're doing just to get something you have a contractual right to, before you can get the simplest thing done...

...or a company where you call a secretary who knows what you're talking about and gets things scheduled for you immediately...

Which company are customers going to prefer?

I'm not buying a company, I"m buying a car...
For instance, I like Tesla more than Ford, but Tesla didn't have the type of car we needed...

Also, most people don't need (or expect to need) to deal with Tesla, so non-factor.

 

[AWDtsla]

No you can not. You might be able to create a cohesive computer system but you won’t ever be able to solve the human element of communication, follow through, attention to detail et al... that is still required.

You only need internal communication when the support systems/technical processes are non-functional. It should be the exception, not the rule.

 

[neroden]

You only need internal communication when the support systems/technical processes are non-functional. It should be the exception, not the rule.

The computer systems are basically supposed to automate internal communications. Amazingly, they seem nearly as bad as the "in person" internal communications.

 

[Krugerrand]

Actually, it often does put them out of business. Depends on whether they have a competent competitor. (Tesla doesn't yet.)

No, actually it doesn’t. I don’t have time to write an essay and give you hundreds of exampkes in this moment. But things that tend to put companies out of business are;

Crappy products
Mismanaging of finances
Larger companies (box stores) undercutting smaller companies
Economic downturns

I'll be pretty f***ing blunt here. When a customer is given the choice between the company where you have to spend an hour on the phone being kicked between different people none of whom know what they're doing just to get something you have a contractual right to, before you can get the simplest thing done...

...or a company where you call a secretary who knows what you're talking about and gets things scheduled for you immediately...

Which company are customers going to prefer?

That may be the choice YOU make. Your reality is not everyone elses. I’ve stayed with businesses that can be very difficult to deal with for a variety of reasons.

Keeping in context here, there is only one Tesla. There will only ever be one Tesla as long as there is only ever one Elon.

 

[Rarity]

Musk wasn't the one who set the culture at Tesla at the beginning, which is why you see the difference in cohesion between SpaceX and Tesla. Musk is doing his best to set the culture at Tesla. He is a master at it. But he is working from behind, given the extreme growth rates constantly breaking the company.

The timeline as I understand it. SpaceX gave its ERM system to Tesla, which extended it and made it its own. Then, several years ago, Tesla's ERM team left to form an ERM startup. And we are left with the consequences.