Hacker News Reader

Minor nitpick, the language is called Malayalam, not "Malalayam". It's easy to remember, because it is the only language in the world which reads the same from back to front ;)

> Therefore a lot of undocumented knowledge gets lost which inevitably causes problems when doing maintenance or solving outages. One week you are working with a brilliant person, the next week they are gone.

I had the same experience with outsourced work to India. Like every place, they have many "rockstar" developers, but these guys do not stay in the same company for too long, an even if they do they don't stay in the same hands-on job for too long and as soon as they move up they will not code anymore themselves, because it is below their pay grade. In India particularly hierarchy plays a huge part in culture and once someone is at a level where coding is seen as "too low" there is nothing you can do to get them continue their job. A company I once worked for even offered a pay rise to their Indian rockstar devs, but wanted them to keep coding and they rather refused the pay rise than doing the "low" work again.

> Thirdly, the sales people at many outsourcing companies are brilliant. They promise the world, then for the first three months keep the clients happy. Then there is always a slump in service / performance as the "dream team", the sales people use to wow the clients move onto the next company they are trying to woo.

Same experience.

I think it's better to hire developers directly and employ them rather than hire developers through a big agency. You won't avoid the staff turnover overall, but it will happen at a much slower frequency.

The “coding seen as too low” mentality is slowly changing. Most of the very well paid employees and even many of the founders of newly found companies are known to be good programmers. People who look down on code are (rightly) missing out

There are two main reasons to outsource: reduce cost & focus on core business. By outsourcing a developer position, a company may not have to be engaged in most of the process of hiring someone else in that geography (It has a cost), there is also the inherit admin cost associated with maintaining the pipeline of developers to cover for the high turn-over, the need of finance/HR/ops professionals, etc from that geography.

In essence, unless a company intends to have major offices in the region and that it may be interested in becoming an outsourcing agency in the future, it may not make sense business wise.

I also experienced the high-turnover thing.

This was probably about 8 or 9 years ago now but the guys from India I workeded with at a BigCo (i.e. both me and them were at the same BigCo) were saying that as soon as they get some real experience of working hands-on and directly with engineers from the UK (and I presume the US as well) then that is seen as a huge thing in terms of their career and they'd be fools not to cash in on that. So you'd work with someone for 6 months, maybe 9, then they'd move to a competitor for a big promotion and a 50% pay rise for their proven skills in working with people from the UK.

You can't blame them for that. Sometimes you have got to play the system to get ahead, because sure as hell the system won't care for you :)

Been my experience as well. I had an offshore contact in India who I trained on our software. I watched the suffix of his email address change 3 times over 18 months. And each time was a significant pay increase.

Everything you say is right and true for out-sourcing. However, I think the better point is that companies like this treat their IT platform as a cost center. They need to invest in that just as much as new planes and amenities for passengers. This is the consequence of not doing so and having 30 year old systems still in the critical path. They have all built a spectacular mountain of technical debt. This is largely true for banks and hospitals and many other industries.

Even though whatever you said is true of offshoring, don't underestimate the incompetence of the in-house talent.

IBM calling.

We had IT support outsourced for all (almost. some branch with government contracts was able to resist) our F500 companies offices. Local IT was let go and replaced by freelancers that have been hired by a company IBM outsourced local IT to.

So we're having: IBM (India) -> outsourcing company (local country) -> freelancers

The local freelancers obviously have no idea, experience or even motivation to stay at the job following phone orders from the operators in India.

This works if you have to do minor things like install a program on a client PC or similar. It becomes risky and ridiculous at server level.

In the end not even basic support worked. People we expensively recruited left because the laptop they've been provided did not work with the company network months later. System outages stretching throughout the continents or even worldwide. Customer satisfaction falling. Angry employees.

The guy who initiated the deal left the company. Probably with a nice bonus from IBM and management is stuck with the situation instead of dropping the contract comes up with useless justifications that are probably delivered to them by IBM.

In the end what might have looked good on that excel sheet ( IT staff salaries vs. IBM contract), causes costs that are not easily put into numbers (we had an account where we could charge time that was wasted due to IT issues. It was closed within a few weeks).

I agree with most of your points. However, these two stuck out to me:

> Fourthly, when the offshoring occurs, the transmission of knowledge sharing is very hard. The local people who are being let go won't be too facilitating in this documentation as well which is understandable.

This sounds like a setting expectations issue. Developers are expected to write code, write tests, and troubleshoot issues. Why not also expect them to produce written documentation? Every job I've ever applied for states "must be proficient in written / oral communication" in the job requirements.

Just be clear about your expectations up front (when you are signing the contract) and this will not be an issue.

If it does become an issue, sue them for breaching the contract. It's the same way any other B2B sale would work - if the company I hire to clean my offices uses a toxic cleaning product, I'm going to sue them for creating a hazardous workplace.

> Finally, just the geographical aspect. Teleconferencing, skyping etc. really adds an overhead and slows down productivity.

I strongly disagree with this. It sounds like a company culture problem. Plenty of companies are fully remote, and many companies have remote workers. Skype for Business, Microsoft Teams, and Slack are all really great at allowing people to communicate.

If anything, it might be a lack of training on how people can use these tools.

Now geographically it does make a difference. You need to be more deliberate about planning your meetings - someone in EST needs to schedule meetings with someone in India in the afternoon only. They can't just schedule a meeting whenever they please.

>This sounds like a setting expectations issue. Developers are expected to write code, write tests, and troubleshoot issues. Why not also expect them to produce written documentation?

They often do, when sysadmins say "There is no documentation" doesn't mean that there is literally no documentation it's often the documentation looks like this: "Set up a cron job and have it fire every few minutes" instead of "login to the server located at 192.168.1.25 via SSH using the administrator credentials located in the password manager and type chrontab -e..."

I’ve worked with offshore dev teams in india, ukraine (2 locations) and romania.

India was definitely the worst. I don’t think we ever reached a point where the work lost to that team (hiring, training, coordinating, spoonfeeding, mess cleaning) was outweighed by actual productivity from them. In other words, I had the feeling that they should have been paying us instead of us paying them.

Ukraine and romania were different. Still many problems inevitable to outsourcing, but at least the ROI made some amount of sense.

Still, I don’t think offshoring is generally worth it. I saw firsthand how the CEO was looking at only half of the numbers (wages) while they weren’t factoring in losses caused by offshoring. I tried to argue this with him at one point, and his argument basically boiled down to: all these big corporations do it, and they can’t all be wrong. I figure all the CEO’s think that.

I was looking for this comment, where someone just blame Indian IT teams without any factual data that this is caused by the Indian IT teams. Just waiting for an opportunity to pour personal hatred and resentment. Just like I expected someone will blame indian IT teams without any actual data.

Agree to lot of your points. Insipite of all these risks, a lot of management teams find value with outsourcing. I guess they feel risks are every where around.

Things are cheaper and easier to do in India. Try hiring 200 devs in Craigland and then come back and give us your observations.

Delays and cockups are just a cost of doing business. If it gets too bad trust me people get fired and contractors change or companies go out of business. Customers don't know how good they have it, because of how competitive the environment is.

I do agree with some of your points but looks like outsourcing industry is the favorite scapegoat these days for any failure. From other anecdotes in this comment section, various other reason are being thrown around for BA repeatedly having IT failure. So without any proper analysis it would be premature to conclude one way or the other.

It doesn't specifically have to be outsourcing. Just the massive change associated with it. Take something mediocre, and try to run it somewhere else, and it's worse. Similar outcome to moving it across the country with a different set of "locals".