Uber wants to cut costs by shifting its engineering to India
businessinsider.in
businessinsider.in
Imagine if Uber told it's developers: we're cutting salaries by 40%- but you are now full-remote, and we want you to live somewhere less expensive than SV. Spread out across America!
A lot of folks would quit. But some would see the incredible opportunity. In that situation, I'd move somewhere like Chattanooga TN- average rent at around $1000/month, vs SF's $3000-5000. Food is cheaper too. And being a city, it's far less conservative than you'd think (well, okay, it is TN).
They could cut their engineering costs dramatically while improving the overall lifestyle of their developers.
And given we're soon going to be in a post-pandemic world where tightly-packed open offices don't exist anymore, they'd probably just be ahead of the rest of us by a few months if they did it.
This doesn't even account for the difficulty in hiring an engineer of the same caliber in Chattanooga and all the associated overhead like health insurance, etc.. which are much lower in India.
If you already have experience building a high quality engineering team in India (thus know how to overcome cultural barriers), its pretty much a no-brainer.
People don't love moving, and they usually move for better work opportunities, not to accomodate a pay cut at their current job. Moving to a lower cost of living area also means moving to an area with less opportunities, and not to mention moving away from your social circles and the life you have established somewhere.
The only way I can see this working if uber started hiring bew remote engineers in the US, but asking their current ones to move from SF so they can take a pay cut?
Lesson learned, I won’t be moving again without a relocation package
What did they do?
I'd be curious if this will change as more companies offer full-time remote.
I live (well) outside of Boston. About 20 years ago, there wasn't a single real tech employer in the city. That's changed primarily with Kendall and the Seaport. But there's still a huge amount in Metrowest and some in the northern suburbs going up to southern NH.
I've been lucky that through several jobs I've never had a really bad commute except into Boston for a time--but I didn't have to do it every day and there is a commuter rail.
Close to Nashville, Huntsville (huge aerospace & engineering base has fiber) & Atlanta.
Pretty sure they also have municipal fiber.
That rent difference turns out to be $24000-$48000 a year. If salaries are cut by 40%, it would only make economic sense to move if you earned less than $60000-$120000 a year (not accounting for cheaper food).
Whether the lifestyle differences would be positive or negative is highly dependent on the individual. And that's without getting into other opportunities available if you stay in SF.
Also think how much is an extra hour every day (or however long you commute) worth to you in quality of life.
At this point, people arguing for moving to lower cost of living area should just show the math.
> Also think how much is an extra hour every day (or however long you commute) worth to you in quality of life.
Just live close to the office. You can literally pay thousands of dollars a month and still be better of financially.
And really, how much is it worth? Is it worth $100K a year? Because that's how much less disposable income/savings some of the people with higher salaries would find them selves with in the proposed scenario.
Living in the city near the office also depends on the area. I can get out off my town in ~5min at any time of the day. There's no traffic on my street. I've got a creek and walking paths nearby. I have time to cook at home and save on takeaways and potentially improve my health. I wouldn't exchange this for an extra $100k and daily commute.
Living in a smaller city will be different. Having different hobbies will have an impact. I don't see any way to show this makes sense in general. It will be make sense for some people though.
What I object to is people saying that cheapness of such places will make up for massive salary reductions. It won't.
Also, you cannot just assume that just because the company decided to cut salaries by 40%, everyone will be excited to pack their bags and move to say "Kansas" or "Idaho" or wherever in the world. Not everyone wants to be a nomad. Lot of us like the stability of a home, a city where we live and have made friends/family and many other reasons. If that was the case, I am sure a lot more people would be moving out of NYC/LA/SF already. Not the case.
I say this as a small bootstrapped founder who has team members in 3 countries and we are only 15 or so. Remote teams are NOT easy. Not at all. I personally would love everyone to be able to work together in 1 office and call it a day. Yes, I am all up for flexibility to work remotely if needed from time to time. I just don't want to deal with the hassles I explained in my first paragraph.
Finally, lets say that we do look at it purely from a cost perspective. On an average, a "good enough" engineer in India will always be at least 10x cheaper than a good enough engineer ANYWHERE in America. It is just the facts of how the global world is. Don't compete on cost. You cannot beat that. Compete on other things.
There is also another class of software engineers in India who are as good as SV engineers who work at companies like Ola, Bookmyshow, PayTM, Flipkart etc. building quality products at quite a large scale. This is the kind of engineer who is also hired by FAANG/Microsoft/Uber etc. And these engineers are paid very well in comparison to the body shop guys.
So when Uber says they might shift engineering to India, that doesn't automatically mean quality is going to shit.
1). While Uber has been very Bay Area-heavy, it has had a distributed engineering team for many of its core operations for a very long time. For example, many of its core infrastructure was built in Europe. Cross-timezone collaboration hasn't been the smoothest, but of all the large "FANGish" tech companies, Uber actually has the most experience with non-HQ teams building business-critical systems. Even companies seen as very progressive in terms of eng culture, like Stripe, seem to keep the core business close to HQ (this is my subjective observation, and I'd love to proven wrong here, because I think Stripe, as a whole, is awesome).
2). India is a country with many people and many engineers. If there's one strong believe I can state in this post, it's that there are many incredibly talented engineers who are living in India and are choosing to stay there. Uber has experience in building engineering teams outside of HQ. It's not a sure thing, but I think they can do it.
Happy to answer any other questions or comments.
(I guess as a disclaimer, I have zero financial stake in Uber, but I have some friends who still work at Uber, so that may or may not influence my opinion)
Tangential, but the core product of AWS, EC2, was developed in South Africa and (largely) continues to be today.
Although I would argue that S3 is the core product
Inhouse they have a sophisticated and long recruiting process. When outsourcing they take the first they get (and as a new customer, be assured, you do not get the best engineers but you need to fight to get them).
my company tried to outsource a ancillary product we develop in 2011-2013 timeframe. It wound up being non-viable and we brought it back in-house.
I think many don't see how pernicious turn-around times can be. They continue to build upon each other like compounding interest. It can be a very very hard loop to break out of. The cycle of question -> answer -> further question -> meeting can grind progress to a halt when each iteration takes 12 hours.
This article is not talking about outsourcing but offshoring. Uber still has total control on the people, the project and rewarding the right behavior. It will just happen in India instead of the US.
But you are right.
Could a successful outsourcing project by Uber threaten all SV tech jobs?
Also, in every company there are jobs that are trivial and jobs that require talent. I assume the more trivial jobs like bug fixing will be outsourced, while the rest of the engineers are kept for product design.
Cost of living is a red herring. No one’s going to pay me more if I live in Manhattan rather than the suburbs of Boise if they can avoid it. It’s all about the locational differences. If there’s a thick, liquid market for the product you’re selling, your labour, you’ll get a good price for it. If there isn’t employers have substantially more power to set wages. A lot of the reason engineers in SV get paid more is that they’re just better but the engineers in Boulder are pretty great too and fewer of them are getting those wages, even fewer in Boise. All about the competition for talent.
But it's complicated. For that matter, you could argue that the perks of Bay Area weather and great number of employment options should make the Bay Area attractive even if the salaries weren't great relative to CoL. But there's enough competition among large employers who generate enough value per employee that they can afford to offer top-shelf salaries so they do so to entice the people they want.
To the last point, I don't expect dramatic short term changes. But to the degree that things tilt more in the direction of remote work and small distributed offices, I expect some equalization of salaries at least within the US.
Actually, why not Canada? Salaries are again much lower than Silicon Valley, the timezone is the same, the language is the same (except in Quebec), and it would be great to stop Canadian software engineers from moving to the US in droves due to lack of opportunities at home.
To get a good job in India, you will need more than skills - connections and a limited edition degree.
Any references for that?
Indian engineers making more than 20k-25K USD Per Year are considered top talent in most companies across India. Are there engineers making more ? Yes absolutely. Are they a minority ? Yes, absolutely.
But it feels - to my uninformed biased mind - like more of Uber flailing around while slowly failing.
> Could a successful outsourcing project by Uber threaten all SV tech jobs?
No.
It could be part of some downward pressure at worst.
Outsourcing to India has gone on for 2-3 decades. It's not some new disruptive thing.
I guess what's different here is that Uber is a proper tech company and knows how to identify good engineers to hire. The same cannot be said of non-tech companies that have outsourced to consultancies like Infosys and Wipro.
In my experience working with the India team at (one of) the FAANGs, it seemed like they were offloading the work that people didn't want to do in SV to the Hyderabad office (e.g. testing, maintenance of legacy systems).
But it could also be the case that there are innovative things being built in these India offices that people just assume were built in the US.
This isn't a new thing. This is just a common strategy that many companies take once they are large enough to be able to handle the extra overhead of working across a 12 hour time difference.
From the engineering point of view Uber is done. The self driving automation is probably not possible in the next x years. It has a working successful product probably too many developers now working on non-core projects.
Moving it to India they can keep the same numbers employed reduce costs and probably throw away things created there anyways.
My guess is a key lean operations team will remain in sf. Ensuring the system runs and is updated. Self driving research can be done India.. UI screens will be local.
Or just not done. There's no reason for Uber to be working on self-driving at this point. Door to door fully autonomous driving in relatively dense areas--i.e. where it makes sense for Uber to operate--is decades away. Autonomous highway driving--which will probably arrive earlier (maybe much earlier)--does absolutely nothing for a would-be automated taxi service.
On the other hand India has an excellent education system and highly skilled engineers, and any capital flow away from the five big cities in the world that sinks the vast majority of it can only be a good thing.
Explain in which way is it excellent?
Which part do you think Indian education system does well relatively?
Could you be more elaborate?
I can see that statement meaning wildly different things on HN and it doesn't answer my question.
It's a result of huge population and most institutions being trash so students are concentrated in one place. That is in no way a feature of the education system itself.
What do you think about reservation in those institutions?
1. https://en.wikipedia.org/wiki/Reservation_policy_in_Indian_I...
2. https://theprint.in/india/education/iits-take-womens-quota-t...
3. For 10% poor all catch reservation which will be implemented soon - https://www.hindustantimes.com/india-news/10-reservation-for...
As far as I see, all energy spent railing against the reservation system is better spent improving yourself.
> As far as I see, all energy spent railing against the reservation system is better spent improving yourself.
You have approximately 60% seats reserved, the remaining 40% seats are left for everyone else to compete on equal grounds.
This year old article disagrees.
https://www.businesstoday.in/current/corporate/indian-engine...
Have they seen India's track record on tech innovation? The education system there does not produce very innovative people or engineers. However they are great at following exact instructions so I'm sure that part of engineering team will do well
I certainly don't see a lack of creativity or innovation from immigrants from South Asia working in North America.
I agree that India has not been the best in terms of tech innovation but the situation is changing. The start up ecosystem is booming here and investments are following too.
Startup ecosystem is booming on foreign investment especially from china. Local investment is still dire.
Students from IIT in India are filtered from millions of students. They only take about 10-15k students every year and the rest are in NIT and other relevant institutions (around 50k). About 1 million self selected students participate in jee-mains every year. I only meant to highlight that OP's friend may be biased compared to the common situation here.
Here is the article for chinese investment in top Indian startups - https://www.vccircle.com/here-s-a-snapshot-of-key-chinese-in...
With government trying to shut down foreign investment, I don't see investment opportunities in India increasing.
Most of Agarwals,Jains,Brahmin are mostly businessman.(most indian big startup-enterprises are from this community.ex.Paytm,Reliance,Birla,Flipkart,OYO,OLA)
for example,I am jain(Baniya). I am having masters degree in IT(2013),had offer from US based MDM startup still I chose trading. Even now I spend 20 hours a week in IT work(hobby),from node projects,System drivers to my trading business specific IOT devices and even DIY Electronics project. But I dont find myself in IT business in near future.
My cousin who is also master degree holder in IT, He runs successful family owned supermarket with 15000+skus. His store is more advanced in tech then bigger competitors which I bet will earn him fruit in next 3-4 years.
Its personal choice not lack of innovation or talent.
Indian engineers, at least the ones hired from the top institutes, can certainly innovate if they are provided the right environment.
Things are however changing in India due to the rapidly growing startup ecosystem which eventually lead to good careers and high salaries for software engineers back home. This is giving people less of a reason to move.
Also, India has good enough engineering talent that product quality shouldn't be an issue given good technical leadership. If companies can hire people straight out of bootcamps for entry level jobs, they can sure get better value by hiring someone having a very good CSE education at much lower cost. Even the vast majority of work at FAANGs doesn't require a Dave Cutler or Jeff Dean. For example, majority of engineers at Google are not working on Search, Colossus,BigTable,Spanner,Borg or the most important technologies. Most engineers just need to know how to use BigTable/Spanner, and that is much easier than designing them.
Operational costs have only recently become a concern, after becoming publicly traded.
Depends on the devs and the kind of team they are working in. It's difficult to provide an objective answer to that question.
$10,000/month may not go far in SV, but in parts of India, you can live like a king, literally, with servants and such.
Rent for a studio apartment in India will cost less than $200. I'm not kidding. But a similar space in SV can cost 10 times more of that.
Don't they have their own ridesharing companies? I don't see why they need uber's jobs, that wouldn't bring much wealth to them but to institutional American shareholders mostly.