94 comments

[ 2.8 ms ] story [ 178 ms ] thread
This isn't a technical goal. Why should anyone but Uber's investors care?
I don't understand what your question means. Feels like you can answer that to basically every single thread on HN just changing two words.
It says things about the state of the market and how unreasonable goals communicated. That said, it's not a very meaningful miss because of covid...unless someone wants to tease apart how much was execution and how much was the environment.
Let’s ignore the elephant in the room about canceling the founder CEO, and look at the brain drain claim instead. I could maybe buy that argument if they were facing technical/scaling challenges, but that’s not the case – was there also brain drain of the top marketing and sales talent in Silicon Valley?

I guess there was probably a big exodus of people after the IPO. All those employees were locked into their positions due to 90 day exercise windows. Once the stock was liquid they had no reason to stay anymore. Then there was the whole “Uber is an immoral company” shtick that gave employees yet another reason to leave (to go work for more upstanding companies like Facebook, I’m sure).

Yeah, it’s no surprise Uber is having trouble. Demoralize your work force, make enemies of regulators and drivers, experience a multi-year global pandemic... none of these things are auspicious signs on their own. Taken together, it’s hard to imagine that Uber hasn’t peaked.

The whole issue with them not being remotely profitable no doubt doesn't help things either...
> look at the brain drain claim instead. I could maybe buy that argument if they were facing technical/scaling challenges, but that’s not the case

Eh, yes and no. There's that famous thread about how they rewrote the iOS app and managed to pull it off by what can only be described as a sheer miracle. That was a legit technical/scaling issue. But also, the first round of layoffs under Dara came largely because there were too many teams spending time on projects of dubious ROI (there was even an internal joke about how people would build things first and only then write the RFC for it...) and eng spending was kinda out of control.

Layoffs were extremely demoralizing to be sure, but honestly, employees are over it now and the internal atmosphere is back to business as usual.

Recall that pre-pandemic Uber had projections to reach profitability by EOY 2020. Personally I think it's a bit weird to single out Uber as having difficulty retaining drivers when I see reports of labor shortages across just about every service industry. On the rider side, people complain about prices going up, but again, this is happening across the board due to the inflation spike (and doubly so for Uber due to its supply/demand bound elastic pricing structure). Sooner or later, I think the economy is bound to return to business-as-usual with their bosses shortchanging minimum wage employees and people scrambling to pay for bills flocking to Uber for extra cash.

</two-cents>

What happened is that during the pandemic when fares were at the bottom, drivers went out and got other skills.

That’s the biggest reason for inflation too IMO: the US economy was being held back by a shortage of skilled workers. Now that workers have acquired more skills, low wage jobs have fewer people to hire and the economy is more productive.

If you want to put the brakes on inflation, under the current fed system that means raising interest rates to slow investment so those there are fewer higher-skill jobs being created and more low-skill ones. Under MMT it means raising taxes on the top earners. Someone has to be the loser though.

>What happened is that during the pandemic when fares were at the bottom

I don't know where you tried using Uber during the pandemic, but from many of my attempts to use it in Seattle, Cincinatti, and ATL during the pandemic, I had the opposite experience. Prices were simply off the charts.

Prices seem to have been returning more to normal recently (while still being at least 50% higher than pre-pandemic). But during the last year, it was total hell. A ride that would typically take $12 was closer to $30+ (both Lyft and Uber btw). A ride from Seattle to Redmond, which normally would cost around $27-31, was in the range of $55+.

Maybe I am misunderstanding your comment about fares being at the bottom, but from what I saw myself, the fares during the pandemic in 2020 were at their sky high.

My understanding of the argument is that the way Uber's elastic pricing works means that once supply reaches some critical upper limit, there's so much competition among drivers that the payoff for any individual driver gets too low and they are forced to go elsewhere.

Personally, I don't think the pandemic timelines align well enough to support the idea that this is specific to the pandemic (though I do agree with it being a thing outside of the scope of the pandemic). IMHO, there was a period of time when prices tanked because rider demand cratered but drivers were still relying on Uber as their safety net, but as the pandemic went on, drivers - especially the full-time ones - were forced to find work elsewhere because there was simply no income to be made from Uber rides. This conversely meant that when riders started coming back, the driver supply was simply not there to match the resurgence in demand.

The thing about Uber is that it's roughly equivalent to fast food worker type of work: for most people, it's not meant to be a long term career, but rather just a crutch to try to get onto the next rung in the ladder. If you are indeed going to do this sort of work full time forever even despite a market saturated with drivers, employers like McDonalds have much better propositions, e.g. in terms of job stability.

Ah yeah I meant “fares” as in number of riders. Rates were sky high but total trips were down so much the drivers weren’t making reliable money even at those rates. A lot of people sold their cars because they couldn’t afford the payment without Uber or did some training to get a better career. Nobody wants to drive a car for a living the rest of their life.
Ah, makes much more sense, fully agreed then. The number of drivers was definitely on the low end, which resulted in waiting forever for a ride + increased fare prices.
This sounds like a cool story and a nice idealistic narrative, but do you have any evidence for it? It’s honestly the first I’ve heard of this even being a claim/trend.
Just anecdotal; but the people I know who used to drive for Uber have desk jobs they do from home now. It’s a better quality of life; why would they go back to driving a car?
> there were too many teams spending time on projects of dubious ROI (there was even an internal joke about how people would build things first and only then write the RFC for it...) and eng spending was kinda out of control.

Why do people feel the need to use work as their place for hobby projects and time wasting bs. I don't even care about wasting company money and time. Feel like it's disrespectful of co workers time when people are just off doing what ever dumb thing they feel is absolutely important

I'm not sure about other places, but at Uber specifically, it was a combination of explosive early growth (and subsequently teams finding themselves having to justify their continued employment) combined with the notion that "Uber scale" was really really big and required custom solutions (never mind that AWS scale actually is orders of magnitude bigger)
Because that's literally how you get promoted. Reimplement something in a new sexier language and add that to your promo packet
You seem to be under the impression that companies hire engineers to improve the product. My pet theory is that FANGUL/et al. really just poach as many engineers from the talent pool as possible - it’s cheaper to keep the smartest guys on payroll then inevitably buying a competitor for billions of dollars.
You're forgetting the most important thing about Uber: they disabled the automated braking system in their 'self driven' cars and killed a person.
That was tragic and all, but it's just a blip as far as stock prices go. Aurora acquired Uber's self driving unit in December, so SDV isn't even part of Uber's strategy anymore.
In a sane world the management of the company would be in prison for that.
There were multiple failures. Two automated braking systems were disabled, and they moved from a 2 person testing standard to a 1 person standard. I think there might be others, but those are the big ones I remember. It took a number of circumstances for that collision to occur.
The operator watching TV shows on their phone instead of doing their job was kind of a big part of it as well.
Yes! I mentally sort of rolled it into "going from 2 to 1 operators", because fucking around like on becomes easy and less attractive when you aren't alone, but it absolutely should be stated. Thank you. And I guess, plus the person going across the highway at night. Just, a whole lot of things had to go wrong in this case.
> Let’s ignore the elephant in the room about canceling the founder CEO

I'm not sure I follow this one. How would the mere presence of the founder allow them to ignore economic reality? Like, this is fundamentally a business model problem. If anything, it might have provided an opportunity to dump some of the money-burning efforts (self-driving etc)

I was acknowledging that a mob cancelling a founder CEO and kicking him out of his own company could surely have deleterious downstream effects on workforce morale. It’s an arguable point and opinions probably range from “the CEO is a mostly pointless figurehead” to “everything flows from the CEO.”

You really can’t discuss what might have caused Uber to lose its market advantage, without first addressing this topic. But it’s contentious and so my comment was meant as “let’s assume the firing of the CEO has no impact, so we can move on to considering other variables.”

In very rare cases the CEO may be critically important (maybe Jobs at certain points in Apple's history), or critically deleterious (I think there's an argument that late-stage Ballmer was this), but generally, a CEO is just a CEO; there's very little reason to think ejecting Kalanick harmed Uber significantly (keeping him might have harmed them slightly, but probably not significantly).
I couldn't agree more. Company culture should be one of the top primary metrics used to value an investment. Forget all the business school lingo, all the numbers, and all the bullet points. CEO's buyback stocks, and accountants cook the books all the time. There's always room for fudging with numbers and metrics and how you present them, so all that "math" is really just misdirection. What is presented really doesn't matter, because at the end of the day the goal is always to present the company in good light. No ones gonna say "we had a bad quarter, we don't have a plan, and lost our lead in the marketplace, so please give us more funding".

What can't be faked is the experience their employees - the team so to speak - have working there. And if the team at large doesn't get along, doesn't believe in what they are working for, well, lets just say they aren't gonna win any championships.

But how did they lose uber eats business?

I don't get that at all. You have the restaurants, the drivers, the math gets better the more of all this you have (shorter drives, more stops per trip etc). Something happened there. Covid should have freed up drivers initially for more eats stuff.

Did they change leadership somehow around Eats? Because 2017 ish it felt like eats was the main player. Then a few years later a ghost.

There has to be some behind the scenes story! Crappy quality in terms of service? What happened?

Or did doordash just outspend them somewhere in the mix? Taking a quick look at DASH financials DASH is burning cash on something.

Hard to believe Uber didn't have the capital to fight the fight.

Not this is bay area annecdata - maybe eats is big in Boston or somewhere, but around here it's def not feeling #1 anymore.

It’s a great question, but due to the geographical clustering, very difficult to answer with personal observations and anecdata.

I wonder how much of this question you could answer with semi-public data. Are there any sources you could scrape that would serve as a good proxy for delivery market penetration by the various participants?

Uber Eats gained ground incredibly rapidly on other delivery options in my market, and then basically just as rapidly has lost it to DoorDash in the last couple years. It's very curious - did they just initially light money on fire for market share and now are suffering because DoorDash was willing to do the same to them? Or is there some deeper difference?
Building a market is expensive. If you don't have to recoup that expense, you can outcompete those that do.
I used to use Uber Eats and mostly moved to DoorDash. DoorDash throws gobs and gobs of deals at me. I can order 2-3 times a week and use a coupon every time.

I return to Uber Eats only when they give me 30% off or something.

uber eats orders come out to be much more expensive than door dash, grubhub, etc. idk why this is the case, but this is why I don't usually use uber eats. they do usually have a better selection of restaurants, I'll give them that.
One answer: Dashpass. For more frequent users it was a game changer, no delivery fees from even far off places and very low overall fees. And partner with credit cards to give it away for free and suddenly you have a massive customer base with recurring revenue. A customer base that will order more because suddenly delivery isn’t that expensive
I have the Uber Eats version of that free from my Amex and yet I find myself using DoorDash w/o a subscription more. Uber lost some restaurants I liked, and even with the pass hasn't been substantially cheaper!
I know UE has the Eats Pass-- did they lag behind Doordash quite a bit adding this?

I know that at least for a while the Eats Pass was restricted to your metro- if you buy an Austin Eats Pass it doesn't work in DFW, etc.

> And partner with credit cards to give it away for free

yeah, with dashpass being free for the entirety of the pandemic if you had the top chase card, having a lower mimimum, and lowering fees by more compared to eats pass its really hard not take that VC subsidized delivery.

It’s often significantly cheaper for me to order takeout through DoorDash with all the promos/rebates/points than to call in an order with the restaurant themselves.

I’m sure this isn’t all subsidies and is likely Doordash squeezing the restaurant in some way (eg forcing them to accept a lower margin on doordash orders, not letting them charge differently on their own vs through doordash for the same items, etc) but it’s pretty good for the consumer in the short term.

I’ve also had some insane deals for grocery and convenience store deliveries where the cost of food + delivery inclusive of all fees was less than if I had gone to pick it up myself. If you’re buying frozen or shelf stable food that’s almost like free money.

Yeah ALL delivery services charge consumers AND restaurants for every order. However, DD I've heard is less predatory as compared to UE (They have tiered structure that the restaurant can choose from.

It's not only good for consumers, but also some restaurants as delivery services are a gift for restaurants that don't have much of a in place dining experience to offer.

In my market, Uber Eats was pretty limited in restaurant options from Day 1 and DoorDash blew them away quickly. I haven't checked in recently on UE to see as a whole how it's doing in terms of options. However, I know my favorite restaurant is still just DD.
seems like a seething former exec trying to dunk on current leadership. Reality is that the former leadership team sold a bunch of lies about self-driving cars creating massive profits that never had a chance at working. Valuation would have never been that high if it wasn't for the AI self-driving cars being shilled
AI self-flying cars
The AI pilot has been trained on all known routes and passenger volume statistics. You tell it to take you to Chicago, and it drops you in a field in Iowa.
this is the answer right here.

this is emil michael, the former exec who suggested smearing a reporter critical of uber (https://arstechnica.com/information-technology/2014/11/uber-...).

he was a cancer on the company and his "hey remember me" tweet is just sour grapes. worse, he's wrong; uber was trading above 60 for a while definitely putting their total valuation above 100B & the 2017 benchmark tweet said "within 2 years" making august 10 2021 the second year they had "missed this goal"

seems like it roughly did hit that goal earlier this year before receding down on delta fears
What a weird tweet. Firstly, the "goal" was set for two years after 2017, so how is this the 4th annual reminder? Secondly, according to a random google link [1], Uber market cap has exceeded $100B in the last year. Stocks go up and down, that's in their nature.

[1] https://companiesmarketcap.com/uber/marketcap/

I think none of his claims about the cause are the real one. It's more like: Uber undercut taxis by using VC money to try to buy market share, which worked, but now they have to raise prices and less people want to use it (plus there was a demand shock caused by covid). See this with Airbnb too. They're no longer cheaper than a hotel in many cases and the experience has declined for a variety of reasons.
Was Airbnb ever really subsidized the way that Uber was (i.e. the host gets more than you paid?)
Yes. I was a large host in 2012, and whenever someone would cause trouble, AirBnB would pay for a Palo Alto hotel room to get them out of there. People would go full psycho just to get a free hotel room. I tried hosting again recently; it was an expensive nightmare, but probably better for their margins. Too bad because I had really enjoyed the community in those places.
I'm a host now and roughly 30% goes to AirBnB. There is both a Host Service Fee and a Guest Service Fee. They also take out taxes
And as a guest, they do jack shit when things go wrong. The best you can hope after spending hours on the phone explaining your case over and over with each new person you’re getting is that they’ll refund your money with a “good luck finding another place to sleep for tonight, you’re on your own”

Airbnb ruined so many of our vacations in the last few years that we’ve sworn to never ever use them again. It’s not worth the trouble.

And the prices are completely ridiculous now, on par or more expensive than a bona fide hotel.

The big difference is AirBnb ended up being a winner from covid, at least eventually.
> See this with Airbnb too. They're no longer cheaper than a hotel in many cases and the experience has declined for a variety of reasons.

This is such a common criticism of Airbnb. Maybe I’m an outlier but my use of Airbnb does not compete with hotels at all: e.g. renting an entire pool house with friends, a secluded cabin out in the mountains or an apartment with full kitchen and laundry for an extended stay in a city center.

Anecdotally I view AirBnB as a hotel replacement but like the OP said, I've started going back to hotels because the price/quality just isn't there anymore.
You could rent party houses well before Airbnb existed.

If you needed a full kitchen and laundry, well, that would be the BnB, wouldn't it?

No, BnB means you get breakfast in an inn, often shared with other people you don’t know, in a place that other ppl live in. AirBNB means you have an entire place to yourself.
> AirBNB means you have an entire place to yourself.

No, it doesn't.

You’re right. That’s how I use it but I forgot thats not the only way.
That lack of distinction and plausible deniability is how the AirBnb hosts skirted around laws requiring a hotel license to use your property as a hotel.
> If you needed a full kitchen and laundry, well, that would be the BnB, wouldn't it?

No, that would typically be a “vacation rental” (often, these are also available as timeshares) or an “extended stay hotel” (in-unit facilities are similar, other features and target market tend to differ as the names suggest.) A BnB (which AirBnB’s usually are not) is a “Bed and Breakfast”, which is basically a boutique hotel featuring complimentary full breakfast, usually in a shared-table setting with other guests. They are kind of (in terms of the direction of variation from typical hotels) the opposite direction of vacation rentals with full in-unit laundry and kitchen facilities.

For some historical context, I point you to Airbnb's 2008 homepage. Back when they spelled out "Bed and Breakfast" in their name.

https://twitter.com/WebDesignMuseum/status/14254929851737088...

Yeah, AirBnB definitely originally traded on the idea of the kind of hosting associated with a traditional B&B, but that was pretty typical of the exaggerated “sharing” marketing of regulation-dodging two-sided marketplaces (Uber’s “ridesharing” being the other poster boy for this,)
This initial idea was to ship air mattresses to hosts, that's where the air bit came from.
The first two are outliers I'm pretty sure. All of my very best experiences with AirBnB have been of that kind, for what it's worth; VRBO is just as good for that, or at least, it used to be.

I suspect the last one is still pretty common, but/and AirBnB's meat and potatoes is short rentals in desirable neighborhoods of big cities, mostly tourist destinations.

Which is what hotels are for, and hotels don't jack up the housing market, cause problems for neighbors, and otherwise annoy the natives.

I must confess I've done my fair share of this, I was a 'digital nomad' for almost three years and it comes with the territory. I can confirm that the AirBnB experience for this use case has declined considerably relative to, say, 2014.

Many of my recent hotels stays were in units that have been recently remodeled to be suspiciously apartment-like: multiple rooms, working kitchen, laundry facilities, pet-friendly, etc.

I don't travel enough since the pandemic to know if this is a nation-wide trend, but I wouldn't be surprised if there's a solid market for ABNB-like amenities with chain-hotel consistency.

There are chains that have been doing this for decades already; see TownePlace Suites by Marriott.
In my experience, TownePlace and similar brands tend to be located slightly out of the way, often in large commercial developments. Not always walkable or in the heart of where you want to be.
I absolutely love extended stay hotels.

But also if I'm going to a national park with friends - AirBNB has unmatchable options.

Perhaps that market is too niche to justify its investors goals (I'm sure most travel is still business travels) but AirBNBs fill an amazing niche.

You're thinking of budget hotels, but there's a smaller market of higher end "bed and breakfast" or "inn" style hotels that offer more amenities like an Airbnb but with less regulatory discretion. Those used to be way more expensive than airbnbs but now they're pretty similar. I imagine this is heavily regional though since community-level short term rental legislation is obviously driven by the desires of the specific community advocating for it.
I don't think this is right. They have to raise prices in some markets (e.g. California) to meet new regulations, sure, but some cities were already profitable pre-pandemic without money burning shenanigans.

The single largest thing that threw a wrench in the profitability projections was the pandemic, which caused two huge impacts in the supply/demand graph (the first one down when everyone stopped going out, and the second up now that things are going back to normal-ish, out-of-whack economy notwithstanding.) Uber, being a two-sided economy company, is extremely sensitive to abrupt changes in supply/demand.

There's this persistent notion here that Uber used VC money to lower prices across the board, which is ostensibly true to some extent, but the reality is a bit more nuanced in the sense that Uber operates in a hyper-local fashion, with budgeting strategies catered individually to each city, each with the end goal of eventually stopping the `money => growth` strategy and switching to a self-sustainable model (or if that can't be achieved, then GTFO of there grabbing equity on the way out, as it already has in multiple markets).

I used to think that Uber eats meant that an Uber would be dispatched to pick up my food for me, which lined up with it being more expensive than other available options. This struck me as super wasteful and simultaneously dodgy, so I never used it initially, and have never had a reason to use it even after discovering it is actually a different model.
> This struck me as super wasteful and simultaneously dodgy

Unless you're in an urban area where bike delivery is viable, how would that be wasteful or dodgy?

I guess it's mostly in my head but somehow the thought of a random Uber driver puting my food on their back seat as opposed to some specialized container comes to mind, and I found that very unappetizing.
I'm not sure what you think Doordash and others do...they also just put the bag they receive from the restaurant into their vehicle and deliver it to you :/
What exactly do they do differently? I assumed it was a way to keep Uber drivers "active" when fares aren't available.
I have seen cyclists with backpacks around, just like for door dash, deliveroo etc. I thought that is their mode of delivery. Based on your comment it sounds as if my initial assumption actually correct?
I'm pretty sure they started with the Uber drivers in cars, but in some cities or with those on bikes they offer that as an option, too.

I've never used it so I wouldn't know.

> I used to think that Uber eats meant that an Uber would be dispatched to pick up my food for me ... so I never used it

Wait... that's not how it works? I was under the exact same impression as you until this very moment.

I've never been convinced that you can make this model profitable if you expect to pay the drivers a living wage. People are price sensitive which limits how much you can charge. (This goes for other delivery services too like DoorDash.) If you charge the customer enough to cover all the expenses and pay a decent wage it results in a price high enough to cut into demand. At some point we may find the right marginal pricing model for all this but I have yet to see it.
Clearly the model did not include a living wage for the driver. In fact, the opposite--the model depended on pushing costs as much as possible onto the drivers.
Let's say they have 1M rides per day for an average of $10/ride. That's $10M in revenue per day. If they charged a flat 5%, that would be $500k in revenue per day, or $182.5M per year.

Of course, they actually did about 13M rides per day in 2020, and they're taking a much greater slice than 5%. They should be printing money, not burning it. They only reason they're not insanely profitable is complete and utter mismanagement.

>They only reason they're not insanely profitable is complete and utter mismanagement.

It's hard to take this statement seriously, at least look into the business before making claims that it should be "easy". How did you come to your $10/ride number? Over what distance/time are those rides? How are those rides spaced during the day? How many drivers can you keep on standby to service those rides. Those questions are incredibly important into whether you can achieve a sustainable business that you have handwaved away into "$10/ride".

5% of $10/ride would be $0.50/ride. Maybe they could charge a $0.50/ride flat fee for rides under $10. CC processing fees, retail, are about 3% and $0.30 each, so that can be added on to uber's cut as well.

They did 5B rides in 2020. At $0.50/ride, that's $2.5B. That's enough to pay 1k employees $1M, and still leave $1.5B for infrastructure, business expenses, and profit. It's also a business that requires no on-going capital other than operating costs, it's not like they buying steel and stamping out widgets, it's practically free money.

Right now, their revenue is over $2/ride, so what are they doing with all that cash? Nothing useful.

This very much. Ultimately what Uber (and most other gig economy companies) offers is on-demand servants - and there's a floor to how much this service can sustainably cost.

The trick Uber pulled was inventing something that has a clear (and sustainable) market within the wealthy elite, but convincing investors that the product could be sustainably operated for vast swathes of the middle- and upper-middle class.

In other words, the trick was asserting their TAM was way larger than reasonable measure, and having everyone believe it.

The plan was for drivers to no longer be a part of the model.

Travis's goal for Uber was to achieve self-driving at all costs, and he acknowledged that Uber would inevitably fail as a company if they could not achieve self-driving in order to be competitive on pricing.

His resolve to stay private and focus on the long-term while openly accepting that risk is why Benchmark pushed him out in order to take Uber public.

Last time I tried to get an Uber, the app found me a ride then when he was on his way he cancelled the ride. I go on the app and get another ride, on the way he cancelled the ride...

I called a regular black cab and got picked up in 5 mins. I assume both got a more profitable / closer ride on the way to pick me up, so ditched me and got that one. That shit doesn't fly with black cabs here.

My favorite is when they call me to ask where I'm going and then cancel if it's not within 5 minutes of city center or an airport.

I'm guessing uber is relaxing their requirements just to keep drivers around

Unfortunately, that isn't how Uber works. If someone was on the way and cancelled a ride, that is largely wasted time for the driver. The incentives for the driver are to come pick you up and get you in their car as soon as possible (thats when they start earning per minute, per km money).

What is more likely that occurred is that they either saw where you were going (which is revealed upon the ride request being accepted, they can only see how far away ETA wise you are from them when accepting and recently some passenger rating information, experiments have been done about showing the approximate pay of rides before accepting as well) and decided not to pick up OR they are multi-apping and found a closer ride (again not likely on this second one because repeated cancelling of rides lowers the likelihood you will get future rides of high value - given you are killing the user experience for the rider).

You mentioned black cab which leads me to believe you might be in New York or Europe and yes, service there is definitely better but for MOST of the US, cabs were notoriously fickle and would RARELY show up on time and even if they did would then sometimes just leave once they heard where you were going or their card reader would "malfunction" just as you finish the ride.

> Unfortunately, that isn't how Uber works

Don't you mean fortunately

Or, how about this: the projections have always been absurdely inflated and were effectively predicated on not having to follow any local rules, thus avoiding the time to market and legal overhead that traditional companies face when doing business.

You cannot ignore reality forever. What you can do, as Uber proved, is borrow against the future and playing the VC ponzi game until that game no longer works... and hopefully you've IPO'd so the real cost of your early profits is distributed over the foolish greater population.

I'd give them a pass due to the pandemic. This was a force majeure event.
He mentions DoorDash. The service was launched in 2013. Back then I would have assumed that online food delivery already had dominant players and it would be hard for a startup to get significant traction.

How did DoorDash overcome that and become to the biggest food delivery service?