Hacker News Reader

Comment by b_tterc_p | Hacker News Reader

I asked in the recent thread how Meituan could possibly be affording to subsidize restaurant meals to be significantly below cost at restaurant as that didn’t make any sense. Turns out according to this article that it’s simply that. It doesn’t make sense. They lost $17 Billion in 2018 for a shallow moat around an ugly castle.

The next recession is going to hit hard, and I’m guessing a lot of the gig economy jobs will get a lot worse if not disappear.

View on HN

I've never heard the expression "for a shallow moat around an ugly castle" before but it's a wonderful picture to paint for just this kind of situation. I'm excited to start incorporating it into my conversations.

I had the exact same thought. Love it. Stealing it.

I was proud of that one so you’re very welcome.

Same.

I actually pictured it, for some weird reason

> The next recession is going to hit hard, and I’m guessing a lot of the gig economy jobs will get a lot worse if not disappear.

It'll be a double whammy hit. As the unsustainable businesses collapse and prices rise back to reality, it'll cause an even larger drop in net sales due to consumer's reluctance to pay the full cost of their poke bowls. Combine that with a bunch of gig economy workers having their only income source dry up and you've got a financial mess at a large scale.

Most of the cost of the poke bowl is fixed, not marginal.

The full cost of the poke bowl is a function mostly of rent and wages (which are largely a function of interest rates and rents). If a recession hits, wages and rents will fall, and with them the cost of the poke bowl as well.

The marginal cost of poke bowl is like $3 of fish and $0.10 of rice and veggies.

The narrative is that automation will make all of these services cheap, so the "gig economy" jobs get lost regardless to autonomous vehicles and robotic hamburger makers, for example. Only then, do these companies come out on top -- and quite big if they can float until automated solutions exist. At least, that's the story these types of business tell to acquire funding.

That's essentially the bet with these companies. You're betting that paying the price to (1) establish and (2) maintain market dominance (by accepting the billions in hemorrhaging today), is worth it once these businesses make a very remarkable turnaround when their operating expenses drop off a cliff, thus yielding massive profits.

Like anything, it's a bet. It could pan out, and early investors will be rewarded very well. Or it won't. I don't think it's a done deal either way, or you wouldn't have two camps of thought. Tons of people decreeing ride hailing companies, many others cheering them on. There's valid reasons to both sides.

> Only then, do these companies come out on top

Only the ones that also own the automated vehicle/robotic hamburger flipper technologies—and if it's not one of the on-demand firms that does that key job automation piece, all the existing on-demand firms are worthless in the face of the firm that owns the technology that erases the big cost in their business, and therefore can partner with the whichever is willing to accept the smallest share (or just start it's own) and own the whole market.

It is completely valid to question the value of being first to market. Will anybody care about Uber or Lyft if a new player, Company X, has a fleet of hundreds of thousands of autonomous vehicles in all major U.S. cities? Consumers have shown us just how easy it is to switch to a new app.

I concur that the branding wouldn't matter-- if the existing rideshare companies couldn't strike a deal with makers of autonomous vehicles, they won't be getting a share of the autonomous profits. Of course, these companies are actually developing their own technology. But that is far from their singular focus today.

Quietly, I assume there are folks operating on the thought that market dominance doesn't matter. We don't hear about them today very much, because they don't get much media attention.

Everyone is assuming customers are price sensitive, it could be that the losses are designed to achieve monopolies and drive out competition while establishing a large two sided network and once achieved they use monopoly pricing power to extract value from price insensitive customers. It’s a reasonable thesis based on history.

I just personally thing this class of company will never achieve positive unit economics because people are price sensitive, and the service is a commodity with no meaningful network effect

I’m pretty sure price sensitivity is already well studied and proven in this case given all of the surge pricing they already experiment with

I agree, that's why, despite understanding the investment thesis, as I point out in my last sentence, I am skeptical.

"Theres a lot of people that spend years telling themselves that they're going to work on rockets. These same people build ad tech platforms and never leave. If you want to build rockets, just go build rockets." -- a very paraphrased Peter Thiel.

If these companies are waiting for the next wave of automation they need to build it. Uber and lyft are trying with self driving but they haven't bet the farm on it. If Uber had spent $800 million on autonomous tech and staked the company's entire future on it, maybe they'd be further.

If Tesla gets to level 4 automation, I doubt they'd have trouble whipping up a ride share platform that connects directly with their vehicles.

Google+ begs to differ. Network effects are real. A product needs to be 10x better than a network effect alternative to have a shot at replacing it. If Tesla is first to market by a year or two, maybe. But if Lyft hits the market within a few months; as a lift user, you’ll just stick to that app, why switch.

> A product needs to be 10x better than a network effect alternative to have a shot at replacing it.

This isn't true, sometimes it's just fashion. At its inception Facebook wasn't 10x better than MySpace. Moreover, ride sharing doesn't benefit from the kind of network effects that social network do. I'm stuck on Facebook because that's where many of my friends are. I have no reason at all to care at all which ride-sharing app my friends use. Even old-school email is more sticky in this regard, my Gmail account has a bunch of historical email I want to keep - so it's unlikely I'll ever delete it entirely. What's in my Uber account? Ride history? Why do I care about keeping that?

A think a better parallel is airlines. I have almost no loyalty to any airline, I just choose the flights that go where I want with the least cost/pain. Why would ride-sharing be that different?

Uber’s answer to that is pretty clear and close to what you’re saying: airline style rewards program.

But I think you’re right. The breadth of offerings is too small and the margins are too tight to gain any real loyalty

Loyalty comes from a large pool of drivers. How many times would you jump between random apps, and wait 20 min for an available driver before you just decide to stick with one? I have a lot of friends that travel internationally; they strongly prefer Lyft but have to use Uber because it is available more widely. They all talk about how they could look for a local app...but say 'who cares, Uber is evil, but whatever'.

That's why the play is with somewhere inbetween, with continental apps like Taxify(bolt), Ola, Grab and Didi. Or then web-based ride hailing.

If we only had a meta search engine for taxi rides that enabled you to get the quickest ride from any service! Somehow that's a thing in the airline sector but (for now) not in the on-demand ride business.

Or at least not on the client side. Cab drivers appear to have no problem using several apps at the same time and picking the best option.

Platform liquidity is more powerful than you are giving it credit for. Any two sided marketplace has network effects by definition because the value to buyers increases with the number of sellers as does the opposite.

Airlines are not a two sided marketplace; they own the fleet. I agree Uber is more a commodity than facebook, but almost no product ever has been as sticky as facebook.

Facebook might be sticky, and people have it for the sake of "having" it, but before we all became some Product Manager's KPI to raise "number of likes" and "number of shared posts" -- it was far more useful. Interaction with the platform is, subsequently, way down since I only care to read so many political rants from family I'm not politically in agreement with and "actual photo/footage of me doing <x>" memes. It's so overdone. I never thought I'd hit a point where I'd think, "you know what, I'm not going to touch that political facebook post" because I love to discuss politics-- and here we are.

An emphemeral ride you recieve via a ridesharing app is a commodity. Your social connections are not a commodity. The distinction in where (and where not) network effects applies matters.

Small critique, though: if there are no drivers, you won't ditch Uber of Lyft for the no-name platform that mandates a 60 minute wait before you can get picked up since they have so few drivers.

Perhaps, network effects should be a continuum, with ridesharing at the weaker end, and social networks at the stronger end, when it comes to business relevance.

Thank you. I'm not sure how people don't grasp this.

> The narrative is that automation will make all of these services cheap, so the "gig economy" jobs get lost regardless to autonomous vehicles and robotic hamburger makers, for example. Only then, do these companies come out on top -- and quite big if they can float until automated solutions exist. At least, that's the story these types of business tell to acquire funding.

This narrative never made sense to me. Buying all those autonomous vehicles and burger flippers is an enormous expense and comes with additional logistics complications such as maintenance. Even if autonomous vehicles were perfected tomorrow, transforming a company with little to no capex like Uber into a capex behemoth would take many years, tons of real-estate negotiations, construction, etc. I'm not convinced Uber could make this transition faster than Amazon could build and app and start parking a fleet of cars at its already-built distribution centres.

Why would “gig economy jobs” get worse?

If anything, during a recession, more people will want(need?) to become gig workers. Uber/Lyft are generally supply constrained today. If there’s a surge of supply because people need money they won’t have to pay new driver incentives, which is one the areas that cause them to bleed cash today. Demand side will fall a little, but people will still need cheap ways to get to/from places, maybe especially if their car gets repossessed.

This isn't about supply & demand, it's about profitability and if these companies are viable long-term. There can be no gig economy jobs if there are no gig economy businesses.

Uber and Lyft can be profitable today. They are in their most mature markets like SF and NYC. If they need to get to profitability they scale back and cut certain cities. That’s it.

Which would mean eliminating upwards of 90% of their workforce. Thus not sustainable, and the jobs disappear.

You’re conflating profitability, growth, sustainability etc. you’re also conflating what is bad for Uber vs what is bad for drivers

It is sustainable to do that if by sustainable you mean have a profitable long term business that is cash flow positive

It would likely massively hurt share prices because you’re trading growth for profitability

I'm not conflating anything, you're just mixing subjects. You said that gig jobs would get better or there would be more, and I countered that was not the case since a business cannot sustain that number of people if it's not profitable.

I don’t think you understand how Uber/Lyft think about their drivers. They aren’t employees. In fact, drivers are their customers. There’s no “sustaining” drivers because they aren’t paying their drivers anything...they are just making a commission off of the driver’s own earnings. Uber and Lyft can’t and won’t “fire” drivers since, again, they aren’t employees.

The best way to think of drivers is that they are customers who are licensing Uber/Lyft software for lead generation for their own business. In a recession a seller of lead generation software (Uber/Lyft) is not going to turn customers (drivers) away. It just might not be profitable for you as a customer of lead generation software to buy it if there aren’t leads, but I don’t think anyone knows what will really happen to rideshare demand if the economy tanks. I think it will go down, but it might not go down as much as one might think.

It's not sustainable for the employees but it probably would be good for the business. Looking at Lyft's careers page, I see 7 UX, 11 product, 14 data science, and 40 engineering positions open. That just seems like an absurd amount.

The whole point was that the jobs would disappear -- if Uber and Lyft back out of smaller cities to be profitable, all of those jobs are gone, which is exactly the point of the person you replied to.

Most of those gig jobs would start disappearing

There's no "backing out" if we're talking about individual cities in a country they already operate in (save a few edge cases). That isn't really how Uber/Lyft enter/exit markets. They will just not pay driver incentives for new markets, which is usually the lighter fire for incentivizing new supply (i.e. drivers) to start driving.

There's no job to be "lost." They aren't going to turn you away for a "job" or "fire" you because you literally aren't employed by the company. Think of it as if you are licensing Uber/Lyft software to generate leads for your own business. That is literally how they think about it and what the driver ToS says.

That’s true, there’s no real reason to “leave” a city. But if they need to stop bleeding cash in an expensive city, they’ll need to raise prices during a hurting economy. Doing this at scale is going to severely dent their ability to cover fixed costs long term.

Now we see those cities starting to price in the negative externalities of these businesses. The congestion pricing in NYC is directly tied to the rise in ride sharing. This could change the economics even (especially?) in the profitable cities.

It's ridehailing, ride-sharing is completely different. Ride-sharing REDUCES congestion.

Ridehailing increases congestion if the city's public transport system is terrible.

This is not a fault of ridehailing, it's a symptom of bad infra.

I see this repeated as an article of faith but how profitable? Uber is billions in the hole. If they make a billion in profit every year for 15 years then they will have repaid their investors money at nominal rates. In real terms they will still be down.

How much profit can Uber make if it cuts 'certain' cities?

I don't think Lyft/Uber should be grouped in with the other gig economy players. Taxis have been a thing for a century and Lyft/Uber represent a real improvement over the traditional model. It's effectively a certainty that we will have app summoned taxis 10 years from now. I don't see a good reason to think that it won't be Lyft/Uber providing them.

Uber and Lyft may well survive, but Uber Eats, Postmates, Deliveroo, Handy etc probably won't fare so well. People won't pay to do menial tasks they can do for themselves if they're at all worried about money.

> Why would “gig economy jobs” get worse?

Because...

> If anything, during a recession, more people will want(need?) to become gig workers.

At the same time, the demand for gig services will drop.

So, to the extent possible, things will get worse for the less-scarce workers, but they’ll also get worse for the firms, because they are already often squeezing gig workers as much (or, in some cases, more) than they can legally get away with, so they don't have a cushion to pass on the demand drop to gig workers and see no net harm.

And them the prices will adjust. It's not going to change for the platform, it might actually raise the amount of orders/trips.

Maybe, but I think not. If there’s a recession, it becomes harder to get cash. If it is hard to get cash, you probably can’t afford to keep burning it in hope of building your competitive moat, especially if people are more conservative to the idea that the moat will never happen. So what does Meituan do about its annual $18B deficit? It can try to raise prices and lower wages (maybe you’re right and supply of labor increases), but that’s a fundamental change to its entire business and a huge gap to fill; plus the wages are already near the bottom. If it doesn’t click with consumers quickly... the business just dies- and for a lot of these services the enticement is just the low price right now. A lot of gig economy jobs are premised on bad businesses that only work now because there is a lot of capital to go around. It might be true that at a certain scale it would work, but that doesn’t mean they’re able to get there.

Can Uber and Lyft do better? Maybe. They can’t drop wages much lower either because these people need not only to eat but to pay off vehicle leases. And the more people driving for Uber, the less each person makes due to limited demand for rides (also likely reduced in a recession), so I think there will be a limited surge of new drivers.

It could be that Uber could just drop all of its scale except where profitable, but that throws billions of sunk cost down the drain too and throws a horrible signal to the market about its prospects for the future.

Tl;dr: I believe that companies with deeply red income statements will suffer the most when a recession hits, and it follows that disposable contract workers for these companies will be the most vulnerable.

Drivers are customers of lead generation software licensed by Uber and Lyft. When you think that way, you’ll realize why it doesn’t make sense to think that Uber or Lyft would have any reason or incentive to “fire” drivers during a recession.

You wouldn’t “dispose” of your customers in a recession.

> a shallow moat around an ugly castle.

I hope this ends up in Matt Levine’s Money Stuff segment as part of his Unicorn Enchanted Forest series.

They lost $17 billion in 2018 and I haven't even heard of them.

I am confused. Is Meituan's ticker 3690? This looks like the company referred to and the actual operating loss is 11bn RMB (so about ~$1.5bn)...which is a lot but revenue doubled, and this is kind of a scale-ish business...so?

The number quoted by Bloomberg (quoting from Nikkei) is comprehensive income including the conversion of pre-IPO securities...so not really reflective of operations.

What is kind of staggering is the cumulative losses to equity ~170bn RMB or $25bn. And presumably, there are options and all sorts. Tbh, I am not even sure how this number is correct given the business isn't even ten years old...I don't look at HK companies very often (and I am aware funny stuff happens in HK)...but how the equity account be wrong?

Are you sure?

https://www.bloomberg.com/quote/3690:HK

This says net income was -137B HKD or about 17b usd

Yeah buddy, that isn't right. That is comprehensive income (most of the financial websites just report the simple line items, which are usually right but very wrong when they are misleading).

For HK stocks, you get the press releases/filings from: http://www3.hkexnews.hk/listedco/listconews/advancedsearch/s... - and this is the annual report, http://www3.hkexnews.hk/listedco/listconews/SEHK/2019/0311/L...

As said though, I have no idea if I am looking at the right company (you sometimes find that there is a holding company or a stock with a similar name or something). Pretty sure it is the same Meituan...but maybe not (and if it is, I still don't understand the losses they booked to equity).

I guess Silicon Valley's Sliceline plot line really was accurate

Reply on news.ycombinator.com