You’re act so confident at armchair coaching and gazing into crystal balls.
If you could be transported back to 2012 to run Tesla from 2012 to 2019, what, seriously, would you do? Not kidding about how differently you would behave on twitter or radio interviews than Musk. No dodging the question by saying you’d bet the company reserves on the Cubs in 2016. I’m talking about business strategy for the company. You get all the benefits of 20/20 hindsight to do it “better” than Elon Musk and all the naysayers and negativity. How would you direct Tesla from 2012 until now?
It’s easy to belittle and dismiss the accomplishments of someone else. If you were handed the company, I doubt you could do better.
This would be a very long reply, but I’ll give you a few short ones. Part of the reason I feel so strongly is that I didn’t need 20/20 hindsight, I saw a lot of the problems real-time.
I think what makes most people into passionate Tesla fans is an innovative product company that was trying to disrupt a traditional industry in a lot of ways, some better than others.
The Roadster was mostly a bespoke, artisanal proof-of-concept that went better than expected: battery tech worked, was viable on a scale of 1500 units, Tesla didn’t try to do everything e.g. Lotus bodies and standard L2 charging. A couple small revisions continued to improve the car (albeit no OTA updates like we know now) and many of their competitors were DOA in this timeframe.
During development of the Model S, they were trying to build their own luxury sedan on their own platform derived/evolved from the Roadster. No practical concerns to deal with yet like servicing, distribution, they hedged their bets on both battery swap (lol) and building out a Super Charging network (SF to LA was all that was needed to convince them this was right). They nailed their right target market (essentially California tech dorks with money to burn) and all the tax credits people who could afford these cars didn’t need. Remember those insane Colorado credits? They also had those MB and Toyota partnerships to carry them thorough, and they were making a lot of money in the early years selling carbon offsets. They got a ton of good press from the Model S and it catapulted them forward, but now it’s a long-in-the-tooth legacy that needs a refresh and Musk has basically given up on. Taking over the NUMI plant when GM and Toyota left during the great recession was a good move, but it hasn’t been as well utilitized as it could have been (see: illegally painting Model 3s in tents while more than 1/3 sits empty).
The Model X made sense on paper: we need an SUV. It was supposed to have a high-high % of part and platform reuse, and really didn’t. The Falcon Wing doors were a problem, it’s never been crash tested, and it’s both a unique solution (soccer moms had to have a $173k launch edition with biodefense) while also being it’s version of the Aztec. A refresh could also help here, except it too is by the wayside. Musk has basically admitted they made a bunch of expensive mistakes, which you can infer are due to his ego.
The Model 3 was designed to be assembled 99% by robots which was a disaster, as Musk has shared more than once, and the unwavering dedication in the face of all odds made the car late, more expensive to build, drove out a bunch of great engineers and management that Tesla overpaid to hire, and ended in a great car with mixed workmanship issues across the various builds, that is obscenely expensiveto repair for even the most minor damage. Now that will give birth to a 3 row Model Y (see vapor below) which must use a TARDIS to squeeze adults into that third row from everything I’ve seen.
Big mistakes:
Buying Solar City – their #1 biggest mistake by far hands down, including the Solar Roof that v3 will allegedly actually maybe be for sale. Powerwall is a niche product that has always been a distraction.
Trying to do too much by themselves, too fast, e.g. no dealer network, no service departments, no part sources, no body shops, no charging infrastructure (the hurdle they have best overcome in their history). In any of their platform partnerships they could have worked with MB or Toyota or anyone else to provide parts and service for them. Instead they’re leasing malls and closing/opening/closing stories multiple times in the same week because (surprise) leases are long term liabilities that suck when cutting costs, but you can’t just walk away from them.
The solution to every problem is to do it themselves (which over extends them further): Gigafactories (just the battery kind, could have been partnerships with battery makers), leasing (they did try US Bank first, now they finance themselves, which by all accounts is a nightmake, and the MF is junk-bond expensive), building their own car carriers, Tesla insurance (which was supposed to be its own first party, then became a State Farm white-label, which by all accounts isn’t going well). The insurance wouldn’t be needed if the volume product wasn’t so expensive to repair, had such a high loss ratio, and (I think) Auto Pilot wasn’t such a wild card actuarially.
Vapor products with fantasy launch dates: Solar Roof (Super-Duper-Full-Self-Driving coast to coast we promise in 2016, the Semi, CyberTruck, Model Y, Roadster 2). Continuing to announce launch dates for products that are years late and don’t get better.
So what would I have done:
Deactivate (all of the) Musk’s keycard(s) and email(s), remove EM from management and leadership positions (no board seat), and keep him on as a consultant. He’s the South African Heinrick Fisker, he just didn’t overspend quite so quickly the first time. Obviously this is impossible given current ownership and management but it’s what the company needs to survive. You’re never getting this financial house in order with so much commingling of personal and business funds.
Focus on what makes Tesla/Tesla: innovative products that nobody else can do, but also make them easier to build/repair. Refresh the S (including a shooting break) and X, light touch refresh the 3 to improve (as much as you can) build and repair before any more work on Y is done. I think the Roadster 2 could be more like a convertible version of the S/3 without trying to be an M4 Competition Convertible. They continue to make the same mistakes that their legacy competitors learned from and solved for years ago: common platform architecture, shared parts and components, and reasonable amounts of automation that increase every year.
Autopilot needs another house cleaning (most of the engineers left in the spring) and needs to get religion about what it is and where the roadmap is really going, when.
Kill: Semi (sell any tech or license platform), carrier, Tesla insurance. Flame throwers. Any of those non-core pet projects that generate a laugh, get deposits (that mostly cancel), or distract need to go.
Stop doing about 2/3 of what they are doing now, poorly. Replace those with strategic partnerships or invite the market to solve them (in a way they couldn’t 10 years ago)
Sell off non-core businesses: Solar City, I think you could sell the battery production capacity to someone else, in the shared spaces where they build cars lease-back and continue to do so. In an economy where all the unicorns own so little, Tesla owns so much, and not much amounts to appreciating assets.
Figure out if Tesla makes more sense to be a platform business or a car company. As it stands in 2019, I’m not as certain the answer is “a car company”- it struggles to do so many basic things a car company could do. I wouldn’t be capricious about this, but you could do some test-and-learns to figure that out.
Find a captive to do their leasing/financing at market rates (they don’t have to be subvented but jfc they stink)
Most important, build a cash reserve to maintain optionaility (which they don’t have today) and invest in the things they really must, instead of doing everything and extremely overleveraged. People look at what they do and compare to Amazon, Tesla is not Amazon. They are not getting economies of scope and scale for these investments, and they’re not getting cashflow off them. They’re giant CapEx boondoggles.
Many of the PEs and VCs I’ve worked with want that: they have big pools of money to scale a viable model up 10/100/1000x (more input yields accretive output), this isn’t that. They’re a public company that claims to make most of their money from building and selling cars, so they need to focus there. They can do plenty of R&D, but the tech and platform isn’t as compelling as most people go on about because it’s only used by them and the IP is all open sourced now.
Instead, the Tesla fans and their irrational exuberance just continues to feed Musk’s ego, to continue to make expensive mistakes and make everything all about himself (I heard a great first-hand/in-person example of this last week from someone who worked there for a couple years). They keep rewarding the bad behavior (thinking it’s what got them the best stuff, it isn’t) and the vicious cycle repeats. Musk didn’t found Tesla, but he ran the co-founder CTO out this year, who built an amazing team that did a lot of amazing stuff in the early days (the vast majority are gone, and not just because they cashed out).
btw most of what I recapped here, I’ve said before.
Nice. Very nice. Thank you for taking the time.
I agree with your choice to “Focus on what makes Tesla/Tesla.” - Improve the 3 and use the improvements for the Y. I don’t expect that 3rd row to be for adults. Those are for the kids when the in-laws are in town, or the extra 2 grown people sometimes.
But the thing that makes Tesla, Tesla, are the non-core projects. I see the distractions as a form of advertising. Just not Super Bowl ads or hiring Matthew McConaughey to do some tv spots. Their stunts keep Tesla in the limelight, and keep their own people creative/inspired. “Vapor” products are way more productive than producing Hollywood-styled tv commercials. That cringe-inducing CyberTruck event got all the free coverage and they didn’t have to contract the Screen Actor’s Guild. Model S and X are previous “Vapor” products with fantasy launch dates, yet here we are today with them out on the streets. Delays or not, they’re on the streets and no other EVs can compare.
You did them justice not to dismiss Tesla’s build-out of their supercharger network. That was required for public perception/acceptance of Tesla’s cars.
I’m going to re-read what you wrote again and let it simmer. From what I remember, it was Toyota who gave up on Tesla, wasn’t it? I enjoyed the trip down memory lane you wrote above.
Maybe some of your advice will reach EM’s ears. It’s good stuff.
Some do, some have historically taken engineers off late, critical projects to work on novelties that may be marketing, but aren’t what any of them want to work on (while making an already late project late, the last half of Brook’s law). At the FANG companies, engineers get side projects they want to work on. Being whipped almost daily about how (for instance) late Auto Summon is and then being pulled to work on easter eggs to play on the screen while the car is charging? Demoralizing.
No that was HUGE. But what came along with it was the vapor of battery swap, essentially run as a bake-off. Super Charging was more viable faster, and the company made the right move to build it out, though it was expensive. As the network has aged, Tesla has done a good job of the engineering to make the tech better (e.g. Model 3 fastest charging whose name escapes me) but you have Super Chargers going up and down all over. Thinking about platforms, that’s an area where they could now (for instance) sell the network to a 3rd party to maintain and operate, open up the specs (or make more compatible) with the other charging standards, and just innovate around their fast charging for their platform and license that to the “distributor” (ala Electric America but obviously not those clowns).
I’ve been with Tesla almost since the beginning, and I’ve followed them closely, and I think they have done quite a bit right, but continue to do too much wrong. IIRC Toyota gave up because they (like MB) were licensing essentially the Roadster platform (pre-Model S) old tech for too much money, and the (2nd) Electric Rav4 was over 50k because of that cost. Not long after the partnerships ended Tesla open-sourced the IP. Kind of a kick in the teeth partnership IMO.
Yeah thanks for listening. I’d like to see Musk “retired” and Tesla really take off like gangbusters, but the hardest years look to be ahead and not behind. We’ll see.
I can say with no real disappointment nor distaste that neither of my model 3’s get rated range in the real world.
Never mind that it’s nearly winter but it just goes to show that Tesla is way optimistic if they think 240wh/mi is realistic even with a driver well seasoned in how regenerative braking works and how to best utilize it.
I suppose if I found a set of super light 18’s that fit over the upgraded brakes 240wh/mi is possible with effort for any dual motor variants. Keep in mind I also have a RWD mid-range which is kind of like an SR+ that actually gets rated range.
I interviewed with Tesla for an engineering position about 9 years ago. It was amazing how little the left hand knew what the right hand was doing at the time. I’m glad I went a different route after hearing all these stories.
Uh… the car was on autopilot and due to federal regulations there has to be someone in the vehicle as a fail-safe in case something were to go wrong… which nothing did.
There was also an engineer onboard to monitor the vehicle.
You don’t honestly think any truck, Tesla included, is driving without someone on board anytime soon that distance, on a highway and not a course, do you?
I know someone has to be present in the vehicle to take over as per regulations. Title says it was driverless, which indicates no one is in the car. I get the point is to convey that it’s autonomous driving but doesn’t mean it’s driverless. There’s a driver present with hands off to take control if something goes wrong.
No I don’t. Title says truck was driverless, it wasn’t. As I said above, driver was present to take control. And that’s exactly what I pointed out in my initial post. You don’t think a car/truck with someone in the driver seat can be driverless, do you?
As an aside, waymo has driverless cars (no trucks yet) on the road with no one in driver’s seat, which…err…makes it driverless.
The person required in vehicle as per regulations sits in the backseat.
Acceleration boost option showed up under my Tesla account this morning. The speculation (based on the website source code) is that for $2k Tesla will unlock the software limit to boost the 0-60 either by 0.5 or 1.2 seconds. This is another upgrade on top of the $7k FSD option that they can sell to existing owners post-purchase via OTA update. So unlike any legacy manufacturer, Tesla has a business model that can make an additional $9k of every existing customer or discount/bundle these upgrades if they need some quick cash. It looks like the stock has also reached 52-week high because of this acceleration upgrade news.
I personally wouldn’t (and in fact did not) pay $7k for FSD that is still at beta stage but if it also lowers my 0-60 to 3.2 seconds and effectively making my car a performance model without the bigger brakes and sport suspension then there goes my $7k. Or if the standalone performance upgrade at $2k gives people 0-60 in the low 3 second range, there will hardly be any existing LR AWD owner not pulling the trigger on that…
Nowhere in the article did it state that a driver had to take over in the truck or not. He was there to take control if needed. Maybe he did, maybe not.
I’m not sure why it matters if the driver is in the back seat, driver seat, passenger seat or trunk. If the truck is driving, the truck is driving regardless of the position of the driver. In theory, the guy in the back seat in your waymo example could hop the seat and take control if needed. How does that make it “driverless,” outside of a few extra seconds to take control vs someone sitting behind the wheel of this truck?
I don’t see the logic behind paying 2k for a max 1.2 second 0-60 time boost unless you’re racing this car. I mean, does it really matter in stop and go traffic that you can get to the next light a few tenths of a second quicker? I mean, don’t get me wrong, I’m all for speed too, but there’s only so much power you can really use on a highway or local traffic.
