The story is always the same with this type of company: take a mature industry that is profitable on a unit basis because it’s boring and unpleasant, then build a narrative around some strategy to make it exciting (giant vending machines!) and get buy-in to spend huge amounts of money in pursuit of the narrative but eventually discover the only way to be profitable is to do what the mature industry players already discovered — but now you’ve got so much debt to service you have to cut even more corners and somehow manage to spend billions on becoming a worse version of what already existed and whatever goodwill you earned is burned.

There’s lots of room for businesses to improve on the boring legacy industries with low margins — like car buying and selling — but it requires careful iteration, it requires taking the established understanding and then building on it. Subsidising the cost of good-but-unprofitable service using investment dollars (Carvana was losing thousands per sale pre-pandemic) doesn’t build a sustainable business unless it’s part of a strategy.

It's ironic that free market types are completely for subsidies when the private industry does it, although public and private subsidies both have issues.

At least government subsidies have a benevolent intent usually. VC subsidies are usually about hiding the true market value of a product to attract customers who otherwise wouldn't have shown interest until it's time to crank the money milker.

People didn't seriously upend their lives to drive for Uber until they retire. The people who were hurt by Uber were people who bought taxi medallions and suddenly found themselves competing with unlicensed taxis.

But on that point, as much as it sucks for them, their monopoly status meant the service was overpriced and didn't innovate. Uber might have been underpriced when it was subsidized, but it's still cheaper than taxis and makes rid hailing easier. Don't forget that even after the reality of things set in, Uber is still a $50B company. There's a real business there.

A more benevolent reading of this is that the VC money is used to make the product more affordable at small scale. Many times products become cheaper/affordable at larger scales. Young companies use the VC $$ help to get them across that inflection point.

This logic breaks down when the product cost doesn’t get lower with more scale. Customer service oriented companies usually fall in this category.

This analogy is strange. Free market types aren't against private investors throwing private money at bad investments nor would they want any regulations against it.

"At least government subsidies have a benevolent intent usually."

Government subsidies aren't known for wasteful spending? They don't hide the real value of a product (like healthcare)?

If the government damages the economy with subsidies, those responsible might have more difficulty getting elected. If a VC firm dumps a juiced bag on S&P index fund buyers while causing a bunch of trouble for taxi companies and first time homebuyers, are any of their clients going to be pissed? My guess is they'll probably congratulate them on a job well done.

They're free to do so, but isn't it cheating when it undercuts competitors while propping up unsustainable businesses in the long run? Subsidies that support predatory pricing don't sound fair.

I don't see why anyone would characterise some people choosing to spend and lose their money as "cheating".

But I want to emphasize, I would have gone to Carvana even if it were considerably more. The trusted, online-only buying experience is just something I'm totally willing to pay for, especially given how extremely awful the normal used-car buying experience is.

At its core, my argument is that there is a sweet spot in which you can get great service from these "boring industry + technology + money" companies but that's because they're actively pursuing good customer service at any cost with no consideration for making money. As soon as the company (or its investors) realise this behaviour is unsustainable, they forget all about their narrative and go all in on making money -- which is when the quality of the service slips.

The amount you pay to Carvana (more or less than you'd pay elsewhere) doesn't have a relationship to the quality of service you will receive, because it's not a "normal" business.

There can be positive consumer outcomes from companies that approach business like Carvana: in the short term, customers get investor-subsidised products, and in the long term, other players in the industry get to learn from consumer reception (e.g: CarMax probably learned about demand for higher quality service from what Carvana demonstrated early on) but Carvana specifically is doomed to failure because it isn't a sustainable company that made a bad decision during the pandemic... it's been unsustainable its entire life.

I am not theorising about their costs, rather, I am referencing their own financials. Per their financials, they must achieve >$4,500 in gross profit per car sold in order to break even on a sale because the cost of providing their service is around $4,500 per car. As far as I know, they have never achieved this and they have only come close during periods where the second hand car market was at its most ridiculous supply-chain induced peaks.

You could start an online second hand car dealership today, with the same online purchasing that Carvana offer, and a better quality of service, and build a great business with great unit economics... if that was your strategy from the first day. Many such companies exist! There are lots of upmarket car dealerships that'll give you excellent service -- and it'll cost you less than $5k over market.

Carvana can't just undo a decade of bad mistakes, they can't just switch from "losing money" to "making money" because every foundational decision about the business was made in the context of "who cares about money, we want growth". You should not think of a company as an implementation of a business model, a company is so much more. Carvana has something like 20,000 employees: at that scale, radical change is basically impossible.

People keep complaining about how flying has become so much more unpleasant.

But other than a small minority, the customers almost always pick lower prices over any improvements in service.

And the same is probably true for buying a used car. Folks may talk about how the experience is poor, but at the end of the day price is what will drive their purchase. So whoever gives the lowest price, which often translates to a worse experience, will succeed in the marketplace.

Unfortunately that summarizes a fairly large fraction of the presentations aimed at VCs.

Sidecar was the "3rd company" that did ride-sharing in my city in 2014-2015, after Lyft and Uber. Somehow they were selling rides from any point in the city to any other point in the city for $1. I suppose it was getting subsidized by VC money; and I took probably $500 worth of rides for a year or so until it shut down.

[0] https://en.wikipedia.org/wiki/Sidecar_(company)

Guess what? Buying cars for too much money and selling them for too little money means they go out of business.

That's literally the opposite of what the GP said:

> The buying experience was simply excellent for me, and I couldn't be happier with the car I bought. I figured I probably could have gotten a comparable deal slightly cheaper somewhere else, but I would have wasted a ton of time

I mean this politely, but did you really read what you were replying to?

Its not, though. Its true the upthread poster focussed on “buying experience” not “price”, but the adverse parts of the typical used car buying experience are a highly evolved optimization for drawing people in with advertised prices while optimizing actual prices without driving buyers off. Carvana’s better buying experience is neither “opposite of” nor even orthogonal to their failure to optimize price, its a direct consequence of them deciding they didn’t need to.

It was a totally hands off experience; I will pay extra to not be talked to by someone. The checkout wizard was nearly as easy as buying something on Amazon. Which car? Enter your bank account number and social security. Agree to this loan rate. Sign over power of attorney so they can get your first year of registration and plates taken care of. They drive up to my apartment building; drop it off; "take it for a spin, I'll wait"; come back 30 minutes later; looks awesome; done.

I've taken that car to the local Audi dealer for service. They upsell you on everything. They pressure you. "Hey, while you're waiting, why not take this new 2022 A4 for a spin, no problem man my treat." "Yeah your car is in great shape; we could give you $15,000 for it right now, that'd take care of your down payment on this new one, its nice isn't it, don't worry about the monthly we can discuss that later" Just stop talking to me. I will pay so much more to not be talked to (and, really, compared to most dealers nowadays; its hard to say if Carvana is even "more expensive").

I feel extremely, EXTREMELY, confident in saying that this attitude has sold Teslas to three people in my bubble. You can start the process to buy a Tesla with Apple Pay. Seriously. Their new delivery process involves ZERO people; you drive to the store, the app tells you the ID of the car which corresponds to a piece of paper hanging in the windshield, you walk around the parking lot looking for it, if there are issues then people get involved; otherwise you drive away. That's it. I am buying a Tesla right now, even though I rather like the Mach E, because they make it so dangerously easy, and I have no clue how to buy a Ford. What's their delivery estimation? Is the price quoted on the site what I'll actually pay? Do I call someone? God, I gotta actually talk to someone? I gotta drive to the dealer? In person? Screw that, Tesla makes it easier, they get my $60,000.

Carvana made mistakes. But their model wasn't some VC abbaration like so many SV companies. They actually struck on something new, in an industry that needed that innovation, and that discovery has since (or maybe, in parallel) been applied by other companies in the industry. They screwed up, but I'll miss them.

The problem with Carvana was that by offering auto loans to buyers as well as taking out loans to buy cars from sellers, they became excessively levered in a time of rising interest rates and falling demand for used cars. On the upswing they massively invested, thinking this was business success and not shocking leverage, and the downswing wiped them out.

Whoever the carvana replacement is going to be, they should subcontract out the business of offering auto loans and they should borrow less to fund acquisition of cars, because the auto business is cyclical and they need to be able to survive the downturns, which requires less rapid growth during the boom. But the overall business model can be fixed, even if Carvana can't.

The pain is that people get pulled into negotiations that mix in a bunch of different things (purchase price, trade-in, financing, add-ons) and it's all a bit more than they can comfortably afford.

I bought an electric car recently and only put up with the minimal haggle of an MSRP cash deal because I was on a sabbatical and had more time than usual. But my partner and I were ready to go for Carvana before my sabbatical or if we couldn't find a dealer willing to stay low-bullshit with us. I ended up calling 6 dealerships and asking my partner to clear her evening just for the one dealership that offered us a low-bullshit deal.

There is a bit of a paperwork dance--also insurance--but, as you say, that's hard to eliminate totally.

Because I want to save money? It literally took about 5 minutes to ask them to add the ~$1500 price of some factory-installed options to my trade-in price and they had a deal.

It wasn't completely frictionless but far and away the main hassle was just car availability. Had I been willing to take a similar vehicle they had on the lot it would have been quicker.

I think Carvana did make mistakes; but I also think the industry as a whole is very deeply suffering right now. And we only hear about Carvana because they're a VC SV darling.

Alternatively, could they hedge their exposure somehow?

My Tesla was ordered online (Apple Pay) and dropped off at my driveway. The only human interaction was the friendly Tesla delivery associate that had a couple papers for me to sign.

Audi, Volvo and Mercedes all had people trying to sell us last years model, adding packages we didn’t want or completely out of inventory on the vehicle we did want.

It’d be great if the Carvana sales model could be adjusted to market pricing, so the customer experience is the same but would be a profitable business.

The Audi, Volvo and Mercedes dealers may have been directing you at sub-optimal cars, but they were available to drive off the lot that day. If you're willing to wait 3 months, you can certainly order the latest Audi or Mercedes model with the exact specs and options you want. You can even pick it up at the factory in Germany and take it for a spin on the Autobahn.

Did you get to inspect the vehicle before you signed the papers? Have heard numerous stories about that, where people either don't get a chance to do so, or when they do find concerns, are pressured by the delivery associate to accept it, "and we'll fix things at the Service Center".

Similarly, my Audi dealer has never tried to run the numbers on trading my car in when I've been in for a service visit, and the only vehicles they've tried to put me in is a loaner vehicle if the service time will be 2 hours or more (and unlike Tesla, they -always- have a loaner for you, even if it's one from the lot).

Really? That would be a 100% show-stopper for me. They absolutely don't need to know such highly private information.

I'll stick with cragislist and paying cash, which is how I've bought every car I've bought in the last ~16 years.

The experience of buying a car at a dealership is a fucking nightmare. It’s awful. It’s the worst customer experience you can have in the United States.

There’s been some serious issues with the used car market fluctuating over COVID but it’s very reductive to now say “oh they were bad the whole time just another unprofitable unicorn”

Carvana has done the same thing, 20 years later.

Car dealerships, for the most part, have not evolved at all or even tried. The experience is still a huge pain in the ass, and they know it.

I'm sure I didn't get close to the best price, but I got better than dealership sticker for the make/model/year, and I lost zero time to the transaction.

Places like Carmax buy a car, mark it up $2k, and sell it making most of their money on financing and warranty/service stuff. They mark vehicles down something like $500 each month for two months then put it right back on the auction if it doesn't sell (so you'll generally get better deals from vehicles that have been on-lot a while). This is why they refuse to negotiate price.

You may get a slightly better price than the dealership, but not by much when you consider every single factor.

If you want the best possible deal, know the car you're looking for and offer a fixed finder's fee to a mom and pop used dealership to buy it for you. A lot of these smaller dealers will give you a good, fixed rate because it's guaranteed money to them without that normal overhead. It may not be strictly allowed (due to NDAs and such), but some may even allow you to help pick out the vehicle to bid on.

The problem is, there's a lot more lemons on the market then there's great deals. Nobody wants the car that doesn't start twice a week or that's been in an accident and has its chassis skewed.

These sites start out by promoting seller karma or something but even the best sellers have lemons to get rid of. At the start they probably just sell them elsewhere to keep their rating high. Eventually as these sites get so big you can't avoid them, the quality will inevitably drop because they capture too big of a market which just includes a lot of crap.

At the same time investors will want to see that exponentially skyrocketing line continuing so they'll be pushing to cut corners left right and center just as things get difficult.

I don't know Carvana as I don't drive much anymore but I assume the same story supplies here at least in some ways.

Initially, you grow like crazy and with great user reviews, as you're able to pick and choose the highest quality items.

Unfortunately, at some point you exhaust the supply at a given quality level, but your valuation mandates continued growth.

So you lower your quality requirements to obtain more supply, which only buys you a bit more time until you hit the next supply limit.

The only winning move seems to be stay private and accept there's a near-term cap on your revenue growth, after which you will grow much more slowly (at the natural supply expansion rate).

Or go multi-brand, and introduce cheaper brands for the lower quality properties, to maintain the goodwill for the higher quality brands. Many hotels do this with many different brands under the same organization that segment their customers on price and quality expectations.

I always thought that Uber/AirBnB etc should be more aggressive in creating more brand separation between their cheaper options and their higher quality options.

But hey, if you feel that attached to a Hampton Inn as opposed to a Home2 or a Hilton Garden Inn I guess it makes sense. It's all the same to me though and seems like just a waste of paperwork and administrative tasks to continue the facade of three brands in the same building.

AirBnB absolutely needs to, given the vast quality range of their current supply.

And at the same time you start burning all the goodwill that make you so popular to begin with.

Thanks for explaining it more clearly, I don't really have a business head.

I couldn’t get a local dealership to give me the sticker price of a car on their website without showing up in person. I expected all the extra fees they throw in to be hidden but all they would post was MSRP and to call for the price. I call and they pretend to not know the car I am talking about and need to go there to see the specific car. This was multiple dealers.

Carvana though showed me the price, shipping if any, taxes, tag, title before I even tried to buy it.

I remember seeing the ads for them hiring delivery drivers and it amounted to "have a pulse? friendly? We'll train you to drive a large commercial vehicle and load/unload cars, wooo!"

If I remember correctly they were banned from car sales in multiple states, in some cases states banning them multiple times: https://www.google.com/search?q=carvana+banned+in+what+state...

Transferring an out of state title in NC hits you with a road tax up to a defined upper limit ($400 I believe is the absolute highest) based on car value and state you transfer from, so they likely did this to avoid paying that sales tax themselves before selling the car. It's shady as hell.

1. What makes you think car dealerships run on thin margins?

I don't know what typical margins are, but the owners seem to do pretty well. This may be despite low margins. But I'm reluctant to accept your claim without proof.

2. What qualifies as low margin? 1%? 10%?

3. It is entirely possible to make billions on low margins.

A dealer that makes $1000 per car can make $1 billion profit by selling 1 million cars. If the average car sells for $15k, a $1k profit would be a < 7% margin. See question 2 about low margins.

A little mom-and-pop lot in a very small town can sell 5-10 cars per month. It doesn't seem unreasonable for a huge national online dealership to sell 1 million cars.

4. What timeframe are we talking? Five years?

Carmax financials say they do more than $5 billion in a single quarter.

As supply costs are so thoroughly controlled, the biggest way to increase margins is increasing prices either directly or through service contracts and/or financing. There is a slight economy of scale to owning fleets of car transports and the like, but the actual competitive advantage of those things over smaller dealerships is pretty minimal and I don't doubt is completely cancelled out by the extra administration overhead smaller dealerships don't have.

Carmax used to mark up cars $2000 over auction price. That isn't pure profit though. They have to pay to transport the cars from the auction to their dealership. They have to pay for the building/lot and administration. They have to pay for any necessary repairs. They have to pay interest on the vehicles as they are leveraged on the vehicles they buy (one reason they resell so fast, but may lose money that must also be made up elsewhere). Most significantly, they have to pay the sales team.

By the time you're done, they probably make more like $100 or less per vehicle directly and the rest indirectly through warranties, service deals, and loans.

can you disrupt used cars given all the logistics of the physical world? maybe, I don't know, but people here report both carmax and carvana having severe business issues, why?

No doubt, because a company that operates at a loss on VC money and passes the savings on to the consumer, can easily provide a great experience/price.

Take advante of those companies while they exist, but be clear that they can't continue for very long. Either VC money runs out and they disappear or they go public and must start making a profit by increasing prices and slashing service.

https://patrickwyman.substack.com/p/american-gentry

They care more about government then literally anyone else in the area, and nobody else has as strong a financial incentive to politically push back on what they want.

regulatory capture + dishonest sleazy sales practices = happy middleman, but suffering consumer and car manufacturer