I quit: Miseries of an Uber driver
salon.comThe one time I ordered an Uber, we spent 25 minutes waiting while the driver drove around a similarly (but still quite distinctly) named street on the other side of town. After several phone calls, she ended up using her own GPS navigator to find us.
So I can corroborate what this article is saying: Uber has absolutely no clue where you are even given both an address and map co-ordinates. Presumably they rely on a terrible reverse geocoding implementation and discard the other information. That's surprising, because years ago I used to work on firmware for navigation devices and I don't recall reverse geocoding ever causing major issues (there were always minor inaccuracies when you round-tripped the data, but you never ended up on the wrong side of town) - I thought it was a solved problem.
For the record, neither Google Maps nor Apple Maps have ever had trouble distinguishing between those two locations (though humans sometimes do), so they broke it all by themselves.
Author here - very interesting to find that someone else had the same strange experience. Street names have no place in GPS navigation! The instance I spoke about was a time when there had been a concert at the Greek Theatre in Berkeley, and I am sure that all available UBER drivers were busy driving people home - hence, even though I was 20 minutes away, I might still have been the closest driver. But the estimation of "3 minutes" combined the GPS path showing me leaving from Piedmont Ave. in Berkeley - when I was on the same-named street in Oakland - makes you doubt their app works at all!
> After waiting 20 minutes, I called the driver, who did not speak any version of English I am familiar with. He claimed to be relatively near my house but was unable to tell me how he was going to get there. I canceled and tried again. This time I got a young woman who also apparently didn’t speak English well.
As someone whose parents speak English with accents, it really makes me cringe when some Americans complain about non-native English speakers in service level jobs. I would hate to have one of my parents deemed less worthy of a job due to their accents. I understand that some jobs benefit greatly from employees that speak English fluently, but I don't think a driver is one of those positions and I doubt many would pay a premium to have a native speaker as a driver.
Maybe I'm overreacting but statements like the one quoted above always come off as slightly xenophobic. It almost makes me throw out the rest of the authors arguments as a rant.
It's not xenophobic to want not to have so much difficulty that it interferes with the service. When you have a hard time communicating directions, it's a problem. Which is what this person experienced. Xenophobia is disliking of people, not disliking the inability to share directions.
It is one big rant - as someone who has ridden with Uber dozens of times in 4 cities I have never had someone who I couldn't understand.
Ideally there would be no need to talk to your driver on the phone in order to find the car. I have only needed to in a tiny portion of my ridesharing trips. But if you do need to, I think it's very reasonable, and not xenophobic, to expect the driver to be able to communicate effectively with an average American (or resident of whatever region you're in).
"Also, considering the company’s huge profits, Uber owes it to the little guys doing their driving to provide much better software..."
Is it typical for reporters to confuse profits with revenue? None of the decks I've seen show profits, I've always assumed that my rides are subsidized by VC money.
> Is it typical for reporters to confuse profits with revenue?
The author isn't a reporter (or, at the very least, she's new to it). Claire Callahan Goodman is a Bay Area native from a large family, whose careers have included Classical Musician, Software Engineer and Mom.[1]
Author here - no, I am not a reporter, but I know the difference between profits, net, VC, and the difference between a privately and publicly-held company. That said, since UBER is privately held, no one can say what they are "worth" (a term which also has many interpretations).
The article published in Salon was a draft I sent them and they ran with it. Once I knew they were interested, I did a lot more research and number crunching, assuming they would want more facts and detail. They didn't change a thing! They published the article in their "Life" section. I think this section is reserved for non-reporters writing about life experiences and that's why they didn't require expert fact-checking. I do agree that no one knows who gets paid what in the upper-echelons of UBER - but I would guess it's a LOT.
With their current gross revenue growth rates, some are speculating that in a couple of years their annual net profits will be more than $1 billion dollars.
> their annual net profits will be more than $1 billion dollars.
Where do you see that from the article you linked? It says So $10 billion of gross revenue would equate to $2 billion of net revenue. To have profits of over $1B on $2B of revenue would mean a profit margin of over 50%!
Why do you think Uber is about to raise funding with ~$40 billion valuation? They don't have profits now because they are investing in growth.
Let me add an slightly alternate (may be off topic) perspective!
In India, the maps are (surprisingly) accurate on the Uber app. The ETA is close to expected and everything's fine. BUT, the driver has no idea on how to use them! It results in the driver calling multiple times to locate where you are. I think Uber must provide a little bit of training to the drivers on the use of navigation system on their app.
I've used Uber since they showed up in Seattle, and was a big proponent of it. Lately, I've been seeing a lot of issues with the estimated eta being off by 20 minutes or so, which has led me to start using other services (Lyft, sidecar, & crown car), all of which seem to do a better job of giving me an ETA which is within a couple of minutes of them showing up.
I have also seen the GPS estimates degrade over time -- which begs the question: shouldn't those estimate be improving over time, unless something shady is going on?
I would guess that GPS experiences like the article mentions are either driver app errors or scaling problems. That is, it sounds like the server is calculating the ETA of the driver based on a stale location. The location would be stale either because the driver app is not updating it correctly or because even when it is updating location correctly, the server is (sometimes) using an older location, probably due to an eventually consistent data store.
When I worked for Flywheel (Uber competitor, same general needs of "figure out the closest driver and provide an ETA"), we'd occasionally see drivers whose phones seemed to get "stuck" in a location, essentially because the location service crashed in the O/S and would stop updating the app (though this was on Android phones and not iOS, and I'd be slightly surprised to see it happen in iOS).
But I think that the scaling issue is the most likely. Location updates come fast and furious, and must be queried quickly. I wouldn't be at all surprised if Uber is using some kind of big scalable highly available database that could in the event of fairly routine network partitions report stale driver locations some of the time, and/or a cache in front of their underlying store that sometimes fails to get invalidated. That seems like it's consistent with the experience of the driver that the problem tends to happen on particular nights (when they're having some kind of underlying problem with, say, communication between components of their data store cluster), and why you might see it getting worse over time (as they scale).
I've seen my iPhone GPS get "stuck" in one place (sometimes in the wrong country!) a couple times since iOS 8. Once it lasted for hours. It seems it's not smart enough to say "well this location is a day old and I've changed tower locations 10 times, maybe it's not valid anymore".
Why can't they just serve the client's location to the driver's device and then have the drivers device handle the estimate?
That would lose scale-dependent ways to improve estimates like current traffic data from other cars, historical traffic and trip-time records, etc
Also won't help with stale GPS location on the phone (which anecdotally has been happening on my N5 more since upgrade to L as well)
My experience with Uber in Seattle hasn't been so hot, but I rarely use taxis or taxi-like services so maybe it's just the luck of the draw. One driver actively campaigned for me to start using Lyft and another decided that downtown to Ballard (as in just outside University Street tunnel station to 15th and Market) needed to go via the Fremont bridge.
It's easier for me to just take Metro and use car2Go to fill in the gaps.
I took an Uber already a couple of times in Seattle and noticed that the Uber GPS seems to take a longer route than necessary in some cases. A former taxi driver told me too that the Uber GPS is not very good.
I'm wondering if they developed their own routing algorithm which would explain why it isn't as good as other navigation systems.
I've had incredibly mixed experiences using Uber in NYC and DC. It used to be great. Now, it seems that every other driver has no idea how to get both to me and then my destination. In DC, the drivers routinely go ten to twenty minutes out of the way. Almost every trip I have to submit a refund request which is always granted, but it's still a pain. In NYC, drivers have yet to figure out one way streets and watching them go up the wrong avenue and circle around my apartment three times has become a past time of mine. After reading this article, it seems like the GPS might be to blame.
I'm not sure how her back to back drives add up to only $10/hr. I drive Uber/Lyft in San Diego, routinely don't get back-to-back rides and have some long stretches of downtime, and usually make $1000 on a 50 hour on-call week (~30h of actual drive time). So that's about $20/hour. If you minus gas (~150/wk) that's a little bit less; I drive a honda insight, which gets slightly worse gas mileage than the prius, and the per hour and per mile in SF is higher than than in San Diego.
> and usually make $1000 on a 50 hour on-call week (~30h of actual drive time). So that's about $20/hour.
Now subtract car payments, insurance, wear and tear, etc.
Here's an example of how this breaks down:
You work 215 hours per month on average (50 * 4.3 weeks) and make $4,300 revenue during that time (or $51,600/year).
- Subtract self-employment tax 15.3% - ~$7900
- Taxable income 37,650 - ~$5,300 in taxes so subtract that
- You put about 70k miles per year on your car assuming an average speed of about 45mph (30 hours * 45mph * 4.3 weeks * 12) - IRS says that's $.56 per mile - ~$39k or basically your entire car + a bit more in one year. This includes the cost of gas.
- The Prius is about a $23k car, financed completely at 3% for 5 years brings you payments to about $400/mo or about $4,800/year
So 51,600 - 7,900 - 5,300 - 39,000 - 4,800 = -5,400 ouch
Most of your cost is in car wear and tear, so let's waive a magic wand and make that go away, it's a Toyota after all. But we can't get rid of gas costs, so let's substitute that in.
A prius gets about 51mpg in the city. It looks like I can get gas for about $2.80/gal in San Diego right now. You'll need about 1.4k gallons per year. Assuming gas prices stay the same, about $3.8k per year in gas. edit I see I'm estimating about half the gas pricing you are, so just double my figure here.
So 51,600 - 7,900 - 5,300 - 4,800 - 3,800 = $29,800 in take home pay
edit or $26k take home when using your gas figures which makes it an even $10/hr
At your self described work hours of 50hr/wk or 2600hr/year, that's $11.50/hr
I assume you wash your car and keep it clean because of passengers, so subtract that out also. I'm going to guess that your take-home pay ends up at around $10/hr if we assume your car takes no wear and tear of any sort driving 70,000 miles a year.
edit I'm not even counting commercial car insurance, which I'm sure you have, or state taxes, inspections, licensing costs, etc. or any other expenses.
edit Taxi insurance appears to run around $7-10k per year.
edit so let's use the new gas figures, you end up with about $21k take home when you subtract $7k of taxi insurance. Or about $8/hr.
Once you deduct state taxes, property taxes etc. I'm sure there's other expenses I'm missing, I have a feeling your take home pay is about on par with California minimum wage.
That's an excellent breakdown, thank you for doing that.
I think Uber's current pricing advantage over standard cabs relies on their driver's not being aware of these costs (especially Taxi insurance). That said, I think they could just raise prices to cover it and still have a great business.
Also, $10/hour is more than I ever made as a shift worker. It sucks, but plenty of people are willing to work for that rate. Thus the OP may be making a moot point.
Yeah. I think in the end he's making about what he could make working minimum wage...and plenty of people do that.
The IRS $.56/mile takes into account estimated gas costs, wear&tear and insurance. You are double dipping.
Also income tax is deducted off of profits not driver revenue.
Yeah you're right. The IRS mileage covers insurance. You can alternately calculate your true expense and deduct that, for example, since he's operating a taxi service, he's probably required to have taxi insurance or risk committing insurance fraud. Since that insurance is much higher than normal car insurance, he probably can deduct more.
it's not a taxi service, and I have one of the insurance companies that is ok with ridesharing. I am sure they have plenty of actuarial data that allows them to set the rates accurately.
I'm not double dipping because I won't be deducting gasoline or insurance expenses as a part of the .56/mi deduction. (I don't keep my gas receipts)
Sure, uber is not a taxi service in the same way McDonald's is not a restaurant.
So. If I were still a postdoc (~40k/year with three years of experience), I would still be paying car payments, insurance. Now, I make those payments and (I think) I can deduct them from my taxes.
Of course comparing to a postdoc is unfair because you're putting in ~80 hours a week, which is certainly below minimum wage, and the postdoc market is totally insane.
Nonetheless, I am definitely making more than minimum wage - for the first time in my life I have savings and investments. Not to mention the fact that I have launched a nonprofit, sinking quite a bit of my own out-of-pocket to get that running. Granted, the take home was much higher at the beginning of the year (avg 1400/wk) than it is now.
CA Minimum wage is $9/hr.
80 hours/wk * 52 weeks * $9 = $37,440 pre-taxes
> Now, I make those payments and (I think) I can deduct them from my taxes.
If you use your vehicle for only business purposes, then yes, you can. If you mix personal and business use, you have to determine the percentage of use, and you can deduce the business percentage.
Since your business involves your car better than 40 hours/week, you should be able to deduct most of the payments.
> Granted, the take home was much higher at the beginning of the year (avg 1400/wk) than it is now.
So wait, do you make 1000/wk now before or after you deduct expenses and taxes or before? I'm assuming that's your pretax/expense earn?
Regardless, if it's working for you and you enjoy it, that's cool. I just hope you're modeling your expenses and taxes correctly and don't get freaked out come tax season.
Most people I know who 1099 get it grossly wrong every year. Which is why it's recommended you do quarterly taxes.
You should be using pre tax numbers for expenses.
He should be using pre-tax numbers for everything - not only is he not deducting any expenses, he's comparing the final after-tax number with the before-tax minimum wage.
Still, assuming his research is accurate at best it comes out at like ~$12 per hour before tax, unless you play russian roulette on the insurance. This of course assumes that it's your full-time job, otherwise the car payments will further reduce the hourly rate.
Sorry, your wrong.
Self employment tax is higher than just social security tax so you need to look at after tax earrings to compare a 1099 job to a w-2 job. Also, more expences can be deducted as an uber driver than a wage slave ex: a percentage of your cell phone bill and car depreciation.
No, I'm assuming he should be bringing home about what the after-tax minimum wage earner brings home for 50 hour work weeks at minimum wage.
But you are right, I should deduct expenses before taxes.
I don't think the end-result is terribly different though. Maybe $1-2/hr
Can you give an example where what I provided was incorrect?
- Car payments are the same
- Gas is the same
- Insurance is the same
- Vehicle wear and tear is the same
I'm sure I made a mistake somewhere or overlooked something...
At the start:
"You work 215 hours per month on average (50 * 4.3 weeks) and make $4,300 revenue during that time (or $51,600/year). - Subtract self-employment tax 15.3% - ~$7900 - Taxable income 37,650 - ~$5,300 in taxes so subtract that"
You don't pay self employment tax on revenue you pay it on profit. So if your paying ex:10,000 in expences for gas then you would pay (51,600 - 10,000) * 15.3% = 6364$ in self employment taxes.
Yup, you're correct. I should deduct the expenses first.
There are other costs besides gas, there's wear and tear on your car, maintenance, interest (if you financed your vehicle, say through one of Uber's subprime lenders), insurance (does your insurance company know that you are using your car for commercial transport?). Do you spend time washing/cleaning the outside and interior of the car? I would at least double the gas cost, so say $300/wk. Do you pay self employment tax? It seems like you are at closer to maybe ~$13/hr or so, with no benefits. A regular (non-exempt) employee would also get overtime if they worked 50hrs.
San Francisco also might have more competition from other drivers.
This is a common mistake people make in hot shot driving also. (People who use 1 ton trucks to haul cars/equipment around). They think great, I'm getting $1.50/mi and fuel is only costing me $0.75/mi, I'm making money. Then you factor in truck/trailer payment, insurance, business related fees(you need that DOT sticker), wear and tear, and cash to buy a new truck every 3-4 years and suddenly you're in the hole big time and all you're doing it racking up milage on your new Dodge.
Should insurance be higher or lower for commercial transport, and if so, why?
Arguably yes. The vehicle is in-use for a greater percentage of the average day increasing the likelihood of being involved in an accident. Also, the vehicle is expected to be carrying more than the owner leading to additional injury claims in court.
Using a non-commercial insurance policy for a commercial-use vehicle is fraud and routinely leads to declined claims.
The vehicle driver might be aware of that and take the risk anyway, but passengers are usually not aware of this and need some protections.
I'm not interested in that, I'm interested in whether or not commercial transport increases insurance costs on average.
Although this question is slightly less interesting because of typical insurance policies being neutral on the number of kilometers driven.
This is easily one of the best contributions I've read in HN in the last month regarding Uber. Nothing beats someone who has actual experience (for better or worse).
Yeah, $10/hour sounds strangely low. Perhaps she's getting a more-frequent dose of the GPS issues than most, or something else, but I think we can take it as given that the median UberX driver -- while not making the $30+/hour that Uber would like you to imagine -- is making comfortably north of $10/hour. If they weren't, they simply wouldn't still be driving for Uber.
She dropped out after had enough experience and data to do the math. It seems like there are a number of people not realistically adding up their true costs.
Regarding ETA times - They were really, really good on the Peninsula in the bay area - over 50 or so rides, and ETA of 5-15 minutes would usually be within 2 minutes of the correct time.
Here in Dubai - they are ludicrously incorrect. I'll see a "8 minute ETA" that turns out to be 35 minutes, more often than not.
This seems to be a very solvable problem. During certain hours, traffic is very slow. I've only been here a month, but I can tell you at 3:30 - 4:30, you need to add about 15-20 minutes if you are going near the interchange at Wafi Mall - That's without even taking into account the various traffic flow indicators that I'm sure Uber has access to.
I gave up on Uber in Dubai. It seems like the maps aren't accurate enough there - or the roads are under too much constant construction. Because it would take drivers 20 minutes to find me.
After presenting my complaints, Uber gave me a nice credit for my troubles and confirmed it was a difficult operating environment due to local conditions.
Consider that Dubai might have different cultural norms as well, which could affect drivers' hurry/accuracy factor.
That's not it - I've taken UberX here about 15 or so times - and the drivers have all been courteous, professional and timely. A lot of them are black car drivers as well, so it's hit and miss when you summon an UberX vehicle, you might just end up with a Lexus a professional driver.
When I see where the cars are, and where I am, I can usually make a guess as to how long they will be based on a months worth of local knowledge - another example is traffic flows to/from Sharjah, the emirate next to Dubai. In the morning all the traffic comes from Sharjah, and in the evening it all goes the return direction. The Exact same trip from Business Bay in Dubai to Al Qusais might take 15 minutes or 60 minutes - but the time is entirely predicable - Uber just doesn't seem to do it here.
The Uber drvers I speak to have generally have had a much more positive experience.
One difference may be point-of-reference. A software engineer has much higher expectations for income, treatment, etc. than blue collar or unskilled workers do. The Radio Shack article a few weeks back is much closer to what happens to most people in the US.
When you talk to taxi drivers, the experience is much, much worse. They have to lease a medallion. The rates vary by market, but typical price might be $150 per day. Because they lease the medallion, they're generally forced to work insane hours. An 8 hour shift is a guaranteed money-losing proposition. Still, if they make less than $150 -- which happens on a bad day, they come out behind. If they have a good day, they'll make a bit over minimum wage. The income is incredibly variable -- a fixed $150/day cost, combined with variable revenue, guarantees that. That makes financial management tough.
It's a really tough, really bad business.
Compared to that, Uber and similar services are a breath of fresh air.
I can see the issue here. Someone says, "Why is this so hard on me when the company shuffling paper is making all the money?" I could see Lyft or Uber having a competitive advantage if they gave equity or something like equity to drivers. Pre-IPO equity may be difficult for regulatory reasons around # of shareholders, but perhaps bonuses based on valuation?
I think these companies want to avoid making the drivers look any more like employees than they already do.
Frankly, it seems to me that they are in fact employees, over which Uber exerts enormous control, to the point of spying on them, monitoring what they post on social media, what they say to customers about the company, supplying talking points, directing them to disregard certain local transportation regulations.
True - it benefits Uber immensely from a tax perspective (and benefits?) for them to be independent consultants. It's also harder for independent consultants to unionize.
Its going to be brutal on Uber (and Lyft) when the IRS recategroizes all of those drivers as employees, and both companies need to cough up the employment taxes they should've been paying. It happened to Fedex (their Ground division) [1].
[1] http://www.forbes.com/sites/robertwood/2014/08/27/fedex-misc...
I'm not a lawyer - is this a done deal? With FedEx they were working full time, right? If someone is driving for both Uber and Lyft, isn't it hard to say they should be an employee of one or the other?
It's a done deal:
http://www.woodllp.com/Publications/Articles/pdf/AlexandervF...
Also, it wasn't the time required from contractors that put them into employee status, but the requirements they were required to follow.
I meant is it a done deal for Uber and Lyft?
Ahh! No no, not at all. It hasn't even come up yet with a court or the IRS.
I'm surprised she complains about the directions. I was under the impression that it used Google Maps. I'm not an Uber driver but I have used the Google Maps app for driving directions and always found it to be pretty good.
It's really bad. I'm in the triangle (NC) and I've had drivers tell me they use Google Maps from their own android as its much better (I think they use Apple Maps)
This probably depends on where you live? I use Google Maps and when I was living in California it was completely awesome. However, having moved to Seattle recently, it definitely is not as reliable as it used to be. This might have to do with the roads here being much more complicated though.
I've noticed that in downtown Seattle, the GPS has some difficulties maintaining a tight fix on your location. Couple that with how close together some of the roads are and you may find yourself teleporting from one avenue to the other as you move toward your destination.
It works great in suburbia, as expected.
It's been explained to me that tall buildings play merry havoc with the GPS, but I haven't been able to test that out at all.
I'm not so sure about the "lousy GPS". In an UberX ride I had the other week, the driver & I were joking about how dodgy GPS can be, and we both agreed to ignore its suggestions... Uber was suggesting we take a tunnel that would take us away from our destination, and it's a tunnel no regular taxi driver has ever used with me when driving to that destination before.
Well, after dealing with evening crowds & inner city construction work that night, we had to take a number of detours to get to the destination. And near the end of our lengthy detour, there was that tunnel that we'd opted not to take. We probably would have saved 10 minutes. The driver immediately apologized: "Maybe the Uber GPS was right after all."
I have learned to stop second-guessing Waze on a similar basis. It has sent me clear around enough traffic snarls and accidents to earn my trust.
You mean its Navigation Software? I'd hate for GPS to become a synonym for route planning.
Former land surveyor and GPS/GIS technician here (now a developer). I have a lot of sympathy for your perspective. I remember when consumer GPS became a thing and getting a little Sheldon Cooper on folks who used "GPS" as a synonym for consumer-grade hardware: "don't you mean GPS navigation?"
But that train left the station a long time ago, and I've stopped arguing with people about it because it's a losing battle. Every normal I know, from my parents to my network of friends, refers to Garmin and TomTom auto units as "GPS." They have unfortunately become identified in the common mind with the technology.
It's been a synonym now for a long time.
That does not mean one can't protest it ;). You need a way to refer to an actual lousy GPS as well (long fix time, inaccurate). Lumping everything together makes it hard to be precise (like using 'begging the question' wrongly somewhere else in these comments ;)).
"protest _against_ it"
Sorry. I'm still fighting a rearguard action against you crazy Americans' perversion of our language a "long time" ago.
(I know. I know. Someone will be along in a minute to tell me that "protest it" was original authentic British English centuries ago.)
"protest _against_ it"
Good point :). As a non-native speaker, I sometimes mess up prepositions... And of course, have a carte blanche to make errors :p.
I don't like it when words I know lose power either, but language changes under stress of modern usage and wrong becomes the new correct. 'Literal' became the new figurative. And so it goes.
Just uninstalled the Uber app this morning. Never used it, and, given the way they treat their employees and their customers, I see there's absolutely no reason to start.
> Having accepted a rider, the driver has no idea of the destination.
Only if you don't enter your destination in the app, which you can totally do.
This is not true, at least according to my limited experience in Prague. The driver will see the destination I put in only after he STARTS the ride, ie after I get in the car (as opposed to seeing it before or after he accepts the ride).
In the US, at least, entering the destination is optional.
Here too, but the driver doesnt see it either way.
Yep, every time I get in an Uber the driver already knows where I'm going - although they apparently don't know until after they've accepted the ride.
I've had only awesome experiences as an Uber rider (in the Boston area and Bay Area) and all the drivers I talked to seem pretty happy (especially ex-taxi drivers). I agree the software is a bit shotty, but I'd bet they are aware of and working out all the kinks.
My understanding is that Uber drivers are always happy about being uber drivers when talking to customers because there is a "secret shopper" program that considers complaining about the company to be a bad thing.
I understand why the company does it, but that's horrific. There was a sandwhich chain (mostly based in the UK) that even had a collective punishment thing - if someone at a shop was found to be insufficiently cheerful, everyone lost on the weekly bonus.
http://www.newrepublic.com/article/112204/pret-manger-when-c...
Yea in my experience they are rarely interested in talking about Uber itself. I always ask how they are finding it and they just say "yea good" or similar and move on.
I've had the opposite experience. Over the Uber rides I have taken, I've always brought up what the driver thinks of the company and most drivers have spoken poorly about it. They usually give the same spiel about how they hate that the rates have come down and that they are not making enough money.
Does Uber use Apple Maps for its drivers? I would assume so, if they are on iOS. That could certainly explain why several comments in this thread mention Uber's direction finding services getting worse. My rather subjective experience with Apple Maps is that they haven't caught up with Google yet.
I remember reading somewhere that they use the Foursquare API for something.
They might use Foursquare API for geolocating points of interest? Citymapper uses that and I assume it's for a reason (cheaper than Google?)