10 ms·

Uber is an example of a company that wants to be a tech company, but (1) really shouldn't (2) kinda lacks the technical leadership to really be one. It feels l

by voz_ 4y ago

Uber is an example of a company that wants to be a tech company, but (1) really shouldn't (2) kinda lacks the technical leadership to really be one. It feels like the right call, around 2017, was to either: (A) double down and build a platform for every kind of physical multisided marketplace (matching as a service, maps+routing+locations+geofences as a service, physical product ops as a service, etc) Where Amazon went from books->everything->AWS, Uber could have gone rides->platform-for-all-the-multi-sided-marketplaces. OR (B) fire every engineer except the 15 or so it takes to maintain the rider and driver apps and a few services, with no new features, port everything over to common cloud companies instead of their DC (shoutout to DCA, the ugly stepbrother of SJC), and just pivot to value extraction and just maintaining it as the simple taxi app it should be. Instead, they did neither, muddling around building random tech as if they had ads/search/social media margins while losing sight of the customers needs. There's a reason DD crushed them in food. Edit: Since the primary response seems to be "You don't understand!" I was there 5 year, saw how the sausage was way over-made.

bogota 4y ago
Im sorry you think you can maintain the rider and driver app with 15 engineers… I really can’t take anything else you say seriously as your knowledge of uber as a business discredited the rest of your thoughts.
caseydm 4y ago
I agree that a small team of engineers can do big things. But 15 engineers to run Uber ride share in maintenance mode? That's insane. I would love to see a breakdown of each engineer's responsibilities in that scenario.
johnebgd 4y ago
Uber crushed themselves in food. As a customer, I would get the wrong food about half the time in NYC even pre-pandemic. I complained and got a refund the first couple of times but then they stopped refunding me when it kept happening. Clearly thought I was fraudulently trying to get meals for free. Thing is, I’ve been a customer since the black car only days. Been Uber VIP for my high rating as a customer. Problem wasn’t with me. If they can’t get quality control done with their restaurants then they don’t deserve my money. I’d rather go direct. I’ve never had a problem go unresolved going direct. In fact, I’ll often use Uber to find a restaurant near me and then call them direct to order to avoid their lousy customer service.
- voz_ 4y ago
  I think your NYC centric anecdote kinda reinforces my point, in an old-man-yells-at-cloud kinda way. The eats usecase was supposed to be around delivery for places that did not already offer it. If you are talking about the eats experience vs in house delivery, that's kind of an edge case...
  - nrmitchi 4y ago
    “Supposed to be around delivery for places that did not already offer it” is beside thee point. It isn’t an excuse to be so terrible across the board (but note this isn’t unique to Uber and DoorDash is pretty terrible as well)
  - edmundsauto 4y ago
    I’m not sure I agree. People in NYC are in for the market - you just have to beat the existing haphazard setup, create a unified payment and support and order taking platform, and do as well as the incumbents in delivery quality.
  - rexf 4y ago
    Disagree. If a restaurant is listed on uber eats, the customer (johnebgd) should expect to get their order right the vast majority of time. Why should it be the customer's problem to figure out the implementation detail (delivery for new places, not in-house)?
- ravenstine 4y ago
  Most restaurants are horrendously incompetent when dealing with delivery apps. I know because I once drove delivery for a few apps for about 6 months. The vast majority of restaurants treat delivery apps as easy side cash and don't take it very seriously. Depending on the restaurant, since the employees don't see you as a real customer, they'll rush things and miss items. They almost never report orders as being ready through the app, which would make it easier for drivers to schedule their drives and get in and out faster. The result is drivers show up and end up waiting because even after 30 minutes the food still isn't ready when they arrive. The tablet the restaurant is using could be completely off or dead but they won't bother signing on to the GrubHub/Doordash website to stop taking orders. Most of the time, restaurant staff doesn't give a fuck, which is stupid because restaurant staff also tends to hate delivery drivers. Why the hell would you not get the manager to sign in to the website and prevent drivers from coming in and wasting your time? On the other side of the coin, these delivery apps also don't give a fuck. I was way above average in terms of the effort I put in as a delivery driver, but no matter what you're gonna get customers blaming you for things you couldn't have prevented. A bunch of times the customer opened the bag at the door (this was before the pandemic), find an item missing, then scowl at me while telling me they're gonna complain to the app. Not once did I ever get reprimanded by an app or receive fewer orders. I kept in touch with a bunch of other delivery drivers, many of which were far lazier than I was and admitted to screwing things up, and none of them every got so much as an email for forgetting drinks too many times or being way too late. The only way you can really get fired from these apps, absent doing something illegal, is if the area the driver is in just doesn't have enough people ordering food to make it worth paying the driver. Oh yeah, I did a little driving for Uber Eats, btw. Most delivery drivers I knew at the time looked at Uber Eats as the "slut" of delivery apps. It paid the worst, the customers were cheapskates, but you also got lots of McDonalds and KFC orders through it, so it was only really good for if you could fit one of those orders in on the way to another delivery. I stopped doing it pretty early on because Uber Eats effectively had no meaningful driver support. At least GrubHub could be reached, even if it wasn't always helpful. Good luck reaching Uber as a driver in 2018. IMO, Uber had no business competing in the food delivery space. I think people overlooked all these issues because we've been in the honeymoon phase of the post-2008 economy with all these "innovative" mobile apps doing amazing things, so we largely overlooked both the bad customer service and ridiculous pricing. Gradually, the rest of the economy will adapt to not rely so much on the likes of Uber unless something changes. And I would say the same thing about a lot of other aspects of the food economy as a whole. Much of it is of underwhelming quality and the customer service tends to suck, and now it's more expensive than it's worth. If you're gonna eat a bunch of carbs and glorified dog meat with some fake cheez product, you might as well do it at home for a third of the price, if not less than that.
ramraj07 4y ago
I mean you’re saying it, and they’re clearly starting to offer it in ny and other places, but perhaps there’s a reason that didn’t work? I can see a million logistical and practical reasons why you couldn’t just make a generic “app for both sides” and let other market owners embrace it. Uber succeeded initially by actually doing the “dirty and unethical” ground work beyond software. I really doubt their tech stack had anything to do with their success.
vladf 4y ago
Wow, strong disagree. Having interviewed a couple of candidates from this space, even for just rideshare pricing for both demand and supply sides is an incredibly sophisticated statistics and economics problem, where appropriate modeling can have long term revenue impact. Even within this vertical it seems you’d need a lot more than 15 engineers.
- hackernewds 4y ago
  It's mind boggling how most people think it's a simple matching FIFO matching algorithm, instead of how mind bogglingly complex making ride-sharing happen at scale is I would know.
  - nanidin 4y ago
    Maybe the world needs a ride share matching game, in the same vein as Elevator Saga[0]? [0] https://play.elevatorsaga.com/ https://play.elevatorsaga.com/
    - smegsicle 4y ago
      or an openttd mod
  - charrondev 4y ago
    Sometimes I wish it were simpler. I’ve had multiple situations where I couldn’t get a ride into my local city recently. The app says there are rides within 10 minutes, shows me a cheap rate and I hit yes. In 10 minutes no matches have been found. In 10 more minutes it finds someone 20 minutes away. 15 minutes later the driver cancels and I’m back to square one. Now the thing is the cost of the ride is a pittance and I would gladly pay double the cost just to get to my destination in a reasonable timeframe. Instead I just call a cab now, request a pickup at a certain hour and they send a driver. FWIW I’m 20 minutes drive from downtown Montreal, so this is not like I’m out in the sticks or anything, and couple years ago it was much easier to find a ride. Unfortunately I don’t think it’s even another ride sharing company taking the drivers, it’s just a lot of drivers pivoted to doing Uber eats instead.
MonkeyMalarky 4y ago
I love the "small team of focused engineers" trope for apps like uber. It's so easy to blow past that number. Website? That's a team. iPhone app? That's a team. Android? Another team. Data collection and A/B testing platform for all the front ends? Another team. Backend? A team. Infrastructure? Another team. AI solutions for problems like delivery time estimation? Another very expensive team. BI for the executives? You guessed it, another team!
- csande17 4y ago
  An underappreciated aspect of computer programming is that maintaining a program often requires less work than writing it in the first place. Once you finish writing the iPhone app, you've finished writing the iPhone app, and you can get by with a much smaller team--possibly just one person--to handle bugfixes and maintenance. Unicorn tech companies generally don't realize this, which is why they all redesign their apps over and over and add poorly-thought-out features all the time. They're still paying for a build-an-entire-iPhone-app-from-scratch-sized team, and all those engineers are desperately looking for stuff to do!
  - iakh 4y ago
    You've never done iPhone development it seems where there's neverending new or deprecated versions of iOS and "kits". I've had clients that try this without a dedicated team, and each time was a nightmare for them(and hence us) where nobody remembered how anything was done
    - csande17 4y ago
      I have done iPhone development, and I can say with confidence that it is less work to take an iOS 17 app and update it to iOS 18 than it is to take nothing and create an iOS 18 app from scratch. You still need a non-zero number of people who know how the iOS app works, but the number you need is smaller.
  - SNosTrAnDbLe 4y ago
    As long as you have good product requirements, creation is actually very simple. You have freedom to choose the technology and the tools. However, The moment that you discover that your AWS bill exceeds your engineering budget, you will have a different view of maintenance. People will now start calling it resilience, scale, optimization, etc. And then one day, the db will need to be partitioned (almost always, if the product is successful) and then the people who created the mess [almost always using a nosql database as it was available ]will run for the exits. You (or the poor unthanked ‘maintenance’ engineer ) will start learning about why distributed systems are hard to maintain.
jollybean 4y ago
No, they have very competent technical people, and, it's not a simple problem to solve. They reached too far, raised to much money, raise too much ire, sang too big a tune. If they had of stayed 'visionary but rational' they'd be doing just fine. And they will probably end up doing just fine in the long run, expectations have to settle down somewhere.
- airbreather 4y ago
  That may be, but they still seem to get continually outsmarted by drivers finding ways to rort the systems.
  - jollybean 4y ago
    That's not an indication of technically incompetency. Everyone has hackers.
notatoad 4y ago
it always makes me happy to see the "tech bros come in here and think that everything we do is simple and easy to replicate" trope every other industry complains about holds true even when applied to the tech industry.
- tzs 4y ago
  Nowhere did voz_ say or imply that Uber was simple or easy to replicate. In fact he didn't say anything all about replicating them. What he is speculating about is a hypothetical where Uber itself stops development and just maintains their existing systems.
  - menzoic 4y ago
    Maintaining a large system like Uber requires redeveloping a large chunk of it constantly to keep up with the growing demand as the app scales. On top of that, Uber is constantly working of features to maintain regulatory compliance And then there is the need for security teams to ward of hacking attempts and keep the system secure over time
    - tzs 4y ago
      Regulatory compliance is the one I suspect would be the hard one when it comes to switching a company like Uber to maintenance mode with greatly reduced developer staff on their apps and backend code. They would have to deal with regulations changing in areas they are already in, and in area they expanded to might have regulations that aren't just a trivial variation of regulations they have already figured out how to handle elsewhere. For dealing with growth, there are two kinds of growth to deal with. One is growth within a given metro area that they already serve. The other is growth into new metro areas. For things like ride scheduling, driver scheduling, routing, and similar they only have to grow their software to handle that in the largest single metro they serve. Multiple metro areas could then be handled by running separate copies of that software in each metro area. That would require an IT staff that would grow as they expanded into different metro areas, but wouldn't increase the number of developers needed to maintain the app and service code. It would be the things that are shared between multiple metro areas that might need a growing developer staff even in maintenance mode. That would be things like the login system and payment processing. But note that these are things that a bazillion other online companies have to deal with too. It's well known how to do them and there are a ton of libraries and frameworks and services to support that. These are things you get working and then rarely have to touch. It's the stuff in the first category, dealing with the details of what Uber has to do that isn't something every other online site with a large nationwide user base has to deal with, where the hard problems for such a business lie. As far as security goes I think the number of developers needed to deal with security goes up much slower than the number of users. It is mostly a function of the amount of code you have, not how many people use it. So, aside from dealing with regulators, I think a company like Uber could if it wanted go into maintenance mode and greatly cut the number of developers (although not down to the 15 voz_ used in his comment), once they have gotten the parts unique to their kind of business to the point that they can handle the largest single metro area they are in or want to be in, and they have a good enough debugged enough feature set that there isn't room for someone else to come up with a better enough feature set to matter. This isn't specific to Uber or other taxi-like services. It is that way for nearly any nationwide business where the actual goods or services provided are essentially local.
howon92 4y ago
I would have nodded if I had read this comment when I was in college…
mym1990 4y ago
Just wondering where the credibility to make statement (2) comes from? This is a pretty big thing to say about a company that has generally been tackling some major technical challenges for quite a few years now. Yes, the economics of the whole thing has changed drastically during their journey, and there have certainly been leadership failures along the way, but they have obviously had technical leadership for quite some time now.
unity1001 4y ago
> OR (B) fire every engineer except the 15 or so it takes to maintain the rider and driver apps and a few services Yeah. Things are * that * easy to do at scale. You need just '15' engineers.
2143 4y ago
> fire every engineer except the 15 or so it takes to maintain the rider and driver apps and a few services, with no new features, port everything over to common cloud companies instead of their DC (shoutout to DCA, the ugly stepbrother of SJC), and just pivot to value extraction and just maintaining it as the simple taxi app it should be. Uber operates in multiple countries, each of which have very different rules. There's millions of rides happening in each of those places. It's difficult to keep everything running with just a few engineers.
bergenty 4y ago
Who is DD and what do you mean they crushed them in food? Uber Eats is king in every major city I’ve been in.
- astrange 4y ago
  DoorDash. No idea who’s winning, though the business seems fundamentally bad since it’s so expensive for everyone involved.
- fatbird 4y ago
  DoorDash. Uber Eats in Vancouver is hot garbage, and my wife and i have switched from Skip to DD because Skip’s quality has plumetted. DD has really captured reliable service in Vancouver.
jorvi 4y ago
> There's a reason DD crushed them in food. Can people please stop making statements true for America as if they are true for the world? In my country we have UberEats, Deliveroo and a few others, but DoorDash doesn’t even exist.
RC_ITR 4y ago
The problem with the “many markets” idea is that outside of taxis and food delivery, there’s often not enough consistent demand for “on-demand” Plumbers and masseurs and dog walkers didn’t build up reservation books because it was hard to make real-time bookings (phones enabled that), they did it because it’s the most efficient way to plan working hours so you’re not sitting around your entire life waiting for work.
mi_lk 4y ago
Yea sure ~15 engs can operate the tech stack of a 50B company. If you worked there for 5 years and still hold these opinions, what you really signal is that you don't know what you are talking about regarding your company's business model and likely just oversimplify how sausage was made with your experience in probably only 1-2 orgs.
Ekaros 4y ago
The problem really is that Uber isn't in a markets that are "tech". Neither is Amazon retail for that side... Taxi and food delivery are horribly thin margin businesses and you really can't change that. Just because someone is middle man there doesn't mean they really are in tech, specially when they are the contact point for both sides. Now if they were selling platform for other people to operate, I could see them being tech...
airbreather 4y ago
The way the app performance has deteriorated I thought they already had sacked most of their engineers - it used to be rock solid, now it glitches all the time for me.
shapefrog 4y ago
I always felt that the business was just a branding excercise to be leader of the pack when self driving taxis became a reality. On May 5, 2020, during the COVID-19 pandemic, Uber announced plans to layoff 3,700 employees, around 14% of its workforce - Ahh I see what you mean, that is a lot of employees
didntknowya 4y ago
Well Uber is logistics/CS company which a tech approach. I would think that makes it a tech company. Their stack is complicated and there are plenty of logistic giants who could've beaten them if only they were as tech focused.
lenkite 4y ago
> (2) kinda lacks the technical leadership to really be one. Really hard disagree. Their software development model is amongst the best in the industry. That needs strong low-level tech leadership.

by voz_ 4y ago

bogota 4y ago

Im sorry you think you can maintain the rider and driver app with 15 engineers… I really can’t take anything else you say seriously as your knowledge of uber as a business discredited the rest of your thoughts.

caseydm 4y ago

I agree that a small team of engineers can do big things. But 15 engineers to run Uber ride share in maintenance mode? That's insane. I would love to see a breakdown of each engineer's responsibilities in that scenario.

johnebgd 4y ago

Uber crushed themselves in food. As a customer, I would get the wrong food about half the time in NYC even pre-pandemic. I complained and got a refund the first couple of times but then they stopped refunding me when it kept happening. Clearly thought I was fraudulently trying to get meals for free. Thing is, I’ve been a customer since the black car only days. Been Uber VIP for my high rating as a customer. Problem wasn’t with me. If they can’t get quality control done with their restaurants then they don’t deserve my money. I’d rather go direct. I’ve never had a problem go unresolved going direct. In fact, I’ll often use Uber to find a restaurant near me and then call them direct to order to avoid their lousy customer service.

voz_ 4y ago
I think your NYC centric anecdote kinda reinforces my point, in an old-man-yells-at-cloud kinda way. The eats usecase was supposed to be around delivery for places that did not already offer it. If you are talking about the eats experience vs in house delivery, that's kind of an edge case...
- nrmitchi 4y ago
  “Supposed to be around delivery for places that did not already offer it” is beside thee point. It isn’t an excuse to be so terrible across the board (but note this isn’t unique to Uber and DoorDash is pretty terrible as well)
- edmundsauto 4y ago
  I’m not sure I agree. People in NYC are in for the market - you just have to beat the existing haphazard setup, create a unified payment and support and order taking platform, and do as well as the incumbents in delivery quality.
- rexf 4y ago
  Disagree. If a restaurant is listed on uber eats, the customer (johnebgd) should expect to get their order right the vast majority of time. Why should it be the customer's problem to figure out the implementation detail (delivery for new places, not in-house)?
ravenstine 4y ago
Most restaurants are horrendously incompetent when dealing with delivery apps. I know because I once drove delivery for a few apps for about 6 months. The vast majority of restaurants treat delivery apps as easy side cash and don't take it very seriously. Depending on the restaurant, since the employees don't see you as a real customer, they'll rush things and miss items. They almost never report orders as being ready through the app, which would make it easier for drivers to schedule their drives and get in and out faster. The result is drivers show up and end up waiting because even after 30 minutes the food still isn't ready when they arrive. The tablet the restaurant is using could be completely off or dead but they won't bother signing on to the GrubHub/Doordash website to stop taking orders. Most of the time, restaurant staff doesn't give a fuck, which is stupid because restaurant staff also tends to hate delivery drivers. Why the hell would you not get the manager to sign in to the website and prevent drivers from coming in and wasting your time? On the other side of the coin, these delivery apps also don't give a fuck. I was way above average in terms of the effort I put in as a delivery driver, but no matter what you're gonna get customers blaming you for things you couldn't have prevented. A bunch of times the customer opened the bag at the door (this was before the pandemic), find an item missing, then scowl at me while telling me they're gonna complain to the app. Not once did I ever get reprimanded by an app or receive fewer orders. I kept in touch with a bunch of other delivery drivers, many of which were far lazier than I was and admitted to screwing things up, and none of them every got so much as an email for forgetting drinks too many times or being way too late. The only way you can really get fired from these apps, absent doing something illegal, is if the area the driver is in just doesn't have enough people ordering food to make it worth paying the driver. Oh yeah, I did a little driving for Uber Eats, btw. Most delivery drivers I knew at the time looked at Uber Eats as the "slut" of delivery apps. It paid the worst, the customers were cheapskates, but you also got lots of McDonalds and KFC orders through it, so it was only really good for if you could fit one of those orders in on the way to another delivery. I stopped doing it pretty early on because Uber Eats effectively had no meaningful driver support. At least GrubHub could be reached, even if it wasn't always helpful. Good luck reaching Uber as a driver in 2018. IMO, Uber had no business competing in the food delivery space. I think people overlooked all these issues because we've been in the honeymoon phase of the post-2008 economy with all these "innovative" mobile apps doing amazing things, so we largely overlooked both the bad customer service and ridiculous pricing. Gradually, the rest of the economy will adapt to not rely so much on the likes of Uber unless something changes. And I would say the same thing about a lot of other aspects of the food economy as a whole. Much of it is of underwhelming quality and the customer service tends to suck, and now it's more expensive than it's worth. If you're gonna eat a bunch of carbs and glorified dog meat with some fake cheez product, you might as well do it at home for a third of the price, if not less than that.

ramraj07 4y ago

I mean you’re saying it, and they’re clearly starting to offer it in ny and other places, but perhaps there’s a reason that didn’t work? I can see a million logistical and practical reasons why you couldn’t just make a generic “app for both sides” and let other market owners embrace it. Uber succeeded initially by actually doing the “dirty and unethical” ground work beyond software. I really doubt their tech stack had anything to do with their success.

vladf 4y ago

Wow, strong disagree. Having interviewed a couple of candidates from this space, even for just rideshare pricing for both demand and supply sides is an incredibly sophisticated statistics and economics problem, where appropriate modeling can have long term revenue impact. Even within this vertical it seems you’d need a lot more than 15 engineers.

hackernewds 4y ago
It's mind boggling how most people think it's a simple matching FIFO matching algorithm, instead of how mind bogglingly complex making ride-sharing happen at scale is I would know.
- nanidin 4y ago
  Maybe the world needs a ride share matching game, in the same vein as Elevator Saga[0]? [0] https://play.elevatorsaga.com/ https://play.elevatorsaga.com/
  - smegsicle 4y ago
    or an openttd mod
- charrondev 4y ago
  Sometimes I wish it were simpler. I’ve had multiple situations where I couldn’t get a ride into my local city recently. The app says there are rides within 10 minutes, shows me a cheap rate and I hit yes. In 10 minutes no matches have been found. In 10 more minutes it finds someone 20 minutes away. 15 minutes later the driver cancels and I’m back to square one. Now the thing is the cost of the ride is a pittance and I would gladly pay double the cost just to get to my destination in a reasonable timeframe. Instead I just call a cab now, request a pickup at a certain hour and they send a driver. FWIW I’m 20 minutes drive from downtown Montreal, so this is not like I’m out in the sticks or anything, and couple years ago it was much easier to find a ride. Unfortunately I don’t think it’s even another ride sharing company taking the drivers, it’s just a lot of drivers pivoted to doing Uber eats instead.

MonkeyMalarky 4y ago

I love the "small team of focused engineers" trope for apps like uber. It's so easy to blow past that number. Website? That's a team. iPhone app? That's a team. Android? Another team. Data collection and A/B testing platform for all the front ends? Another team. Backend? A team. Infrastructure? Another team. AI solutions for problems like delivery time estimation? Another very expensive team. BI for the executives? You guessed it, another team!

csande17 4y ago
An underappreciated aspect of computer programming is that maintaining a program often requires less work than writing it in the first place. Once you finish writing the iPhone app, you've finished writing the iPhone app, and you can get by with a much smaller team--possibly just one person--to handle bugfixes and maintenance. Unicorn tech companies generally don't realize this, which is why they all redesign their apps over and over and add poorly-thought-out features all the time. They're still paying for a build-an-entire-iPhone-app-from-scratch-sized team, and all those engineers are desperately looking for stuff to do!
- iakh 4y ago
  You've never done iPhone development it seems where there's neverending new or deprecated versions of iOS and "kits". I've had clients that try this without a dedicated team, and each time was a nightmare for them(and hence us) where nobody remembered how anything was done
  - csande17 4y ago
    I have done iPhone development, and I can say with confidence that it is less work to take an iOS 17 app and update it to iOS 18 than it is to take nothing and create an iOS 18 app from scratch. You still need a non-zero number of people who know how the iOS app works, but the number you need is smaller.
- SNosTrAnDbLe 4y ago
  As long as you have good product requirements, creation is actually very simple. You have freedom to choose the technology and the tools. However, The moment that you discover that your AWS bill exceeds your engineering budget, you will have a different view of maintenance. People will now start calling it resilience, scale, optimization, etc. And then one day, the db will need to be partitioned (almost always, if the product is successful) and then the people who created the mess [almost always using a nosql database as it was available ]will run for the exits. You (or the poor unthanked ‘maintenance’ engineer ) will start learning about why distributed systems are hard to maintain.

jollybean 4y ago

No, they have very competent technical people, and, it's not a simple problem to solve. They reached too far, raised to much money, raise too much ire, sang too big a tune. If they had of stayed 'visionary but rational' they'd be doing just fine. And they will probably end up doing just fine in the long run, expectations have to settle down somewhere.

airbreather 4y ago
That may be, but they still seem to get continually outsmarted by drivers finding ways to rort the systems.
- jollybean 4y ago
  That's not an indication of technically incompetency. Everyone has hackers.

notatoad 4y ago

it always makes me happy to see the "tech bros come in here and think that everything we do is simple and easy to replicate" trope every other industry complains about holds true even when applied to the tech industry.

tzs 4y ago
Nowhere did voz_ say or imply that Uber was simple or easy to replicate. In fact he didn't say anything all about replicating them. What he is speculating about is a hypothetical where Uber itself stops development and just maintains their existing systems.
- menzoic 4y ago
  Maintaining a large system like Uber requires redeveloping a large chunk of it constantly to keep up with the growing demand as the app scales. On top of that, Uber is constantly working of features to maintain regulatory compliance And then there is the need for security teams to ward of hacking attempts and keep the system secure over time
  - tzs 4y ago
    Regulatory compliance is the one I suspect would be the hard one when it comes to switching a company like Uber to maintenance mode with greatly reduced developer staff on their apps and backend code. They would have to deal with regulations changing in areas they are already in, and in area they expanded to might have regulations that aren't just a trivial variation of regulations they have already figured out how to handle elsewhere. For dealing with growth, there are two kinds of growth to deal with. One is growth within a given metro area that they already serve. The other is growth into new metro areas. For things like ride scheduling, driver scheduling, routing, and similar they only have to grow their software to handle that in the largest single metro they serve. Multiple metro areas could then be handled by running separate copies of that software in each metro area. That would require an IT staff that would grow as they expanded into different metro areas, but wouldn't increase the number of developers needed to maintain the app and service code. It would be the things that are shared between multiple metro areas that might need a growing developer staff even in maintenance mode. That would be things like the login system and payment processing. But note that these are things that a bazillion other online companies have to deal with too. It's well known how to do them and there are a ton of libraries and frameworks and services to support that. These are things you get working and then rarely have to touch. It's the stuff in the first category, dealing with the details of what Uber has to do that isn't something every other online site with a large nationwide user base has to deal with, where the hard problems for such a business lie. As far as security goes I think the number of developers needed to deal with security goes up much slower than the number of users. It is mostly a function of the amount of code you have, not how many people use it. So, aside from dealing with regulators, I think a company like Uber could if it wanted go into maintenance mode and greatly cut the number of developers (although not down to the 15 voz_ used in his comment), once they have gotten the parts unique to their kind of business to the point that they can handle the largest single metro area they are in or want to be in, and they have a good enough debugged enough feature set that there isn't room for someone else to come up with a better enough feature set to matter. This isn't specific to Uber or other taxi-like services. It is that way for nearly any nationwide business where the actual goods or services provided are essentially local.

howon92 4y ago

I would have nodded if I had read this comment when I was in college…

mym1990 4y ago

Just wondering where the credibility to make statement (2) comes from? This is a pretty big thing to say about a company that has generally been tackling some major technical challenges for quite a few years now. Yes, the economics of the whole thing has changed drastically during their journey, and there have certainly been leadership failures along the way, but they have obviously had technical leadership for quite some time now.

unity1001 4y ago

> OR (B) fire every engineer except the 15 or so it takes to maintain the rider and driver apps and a few services Yeah. Things are * that * easy to do at scale. You need just '15' engineers.

2143 4y ago

> fire every engineer except the 15 or so it takes to maintain the rider and driver apps and a few services, with no new features, port everything over to common cloud companies instead of their DC (shoutout to DCA, the ugly stepbrother of SJC), and just pivot to value extraction and just maintaining it as the simple taxi app it should be. Uber operates in multiple countries, each of which have very different rules. There's millions of rides happening in each of those places. It's difficult to keep everything running with just a few engineers.

bergenty 4y ago

Who is DD and what do you mean they crushed them in food? Uber Eats is king in every major city I’ve been in.

astrange 4y ago
DoorDash. No idea who’s winning, though the business seems fundamentally bad since it’s so expensive for everyone involved.
fatbird 4y ago
DoorDash. Uber Eats in Vancouver is hot garbage, and my wife and i have switched from Skip to DD because Skip’s quality has plumetted. DD has really captured reliable service in Vancouver.

jorvi 4y ago

> There's a reason DD crushed them in food. Can people please stop making statements true for America as if they are true for the world? In my country we have UberEats, Deliveroo and a few others, but DoorDash doesn’t even exist.

RC_ITR 4y ago

The problem with the “many markets” idea is that outside of taxis and food delivery, there’s often not enough consistent demand for “on-demand” Plumbers and masseurs and dog walkers didn’t build up reservation books because it was hard to make real-time bookings (phones enabled that), they did it because it’s the most efficient way to plan working hours so you’re not sitting around your entire life waiting for work.

mi_lk 4y ago

Yea sure ~15 engs can operate the tech stack of a 50B company. If you worked there for 5 years and still hold these opinions, what you really signal is that you don't know what you are talking about regarding your company's business model and likely just oversimplify how sausage was made with your experience in probably only 1-2 orgs.

Ekaros 4y ago

The problem really is that Uber isn't in a markets that are "tech". Neither is Amazon retail for that side... Taxi and food delivery are horribly thin margin businesses and you really can't change that. Just because someone is middle man there doesn't mean they really are in tech, specially when they are the contact point for both sides. Now if they were selling platform for other people to operate, I could see them being tech...

airbreather 4y ago

The way the app performance has deteriorated I thought they already had sacked most of their engineers - it used to be rock solid, now it glitches all the time for me.

shapefrog 4y ago

I always felt that the business was just a branding excercise to be leader of the pack when self driving taxis became a reality. On May 5, 2020, during the COVID-19 pandemic, Uber announced plans to layoff 3,700 employees, around 14% of its workforce - Ahh I see what you mean, that is a lot of employees

didntknowya 4y ago

Well Uber is logistics/CS company which a tech approach. I would think that makes it a tech company. Their stack is complicated and there are plenty of logistic giants who could've beaten them if only they were as tech focused.

lenkite 4y ago

> (2) kinda lacks the technical leadership to really be one. Really hard disagree. Their software development model is amongst the best in the industry. That needs strong low-level tech leadership.