I just cannot see why outsourcing this way is sustainable, and it probably isn't. Some are great, and some are absolutely abysmal, an internal joke in the company was that a senior engineer was the same a junior engineer or worse. I had an a senior engineer who couldn't change a string to another string in a code base. I don't even exaggerate I had to prove to my manager that he had no skills at all.
I don't think the issue is really outsourcing in the sense of sending this work overseas, as contracting important functions outside of Boeing. Sending this work to an entirely US firm would also have been a problem given the way their interactions with Boeing worked:
> The coders from HCL were typically designing to specifications set by Boeing. Still, “it was controversial because it was far less efficient than Boeing engineers just writing the code,” Rabin said. Frequently, he recalled, “it took many rounds going back and forth because the code was not done correctly.”
Key attributes here are:
- The engineers at the external firm isn't necessarily incentivized to get it right the first time.
- Boeing still has to have engineers on staff to write the specs, and the evaluate the work product of the contractors _repeatedly_ to see whether it meets the specs.
- If it's wrong, it's really really costly.
In that context, I think it makes way more sense for this stuff to be built entirely by people on staff.
The article doesn't discuss it and I don't know to what degree this is still at play, but for years after the merger with McDonnell Douglas, a lot of the Boeing people were frustrated with how processes were changed to depend on many more outside firms. Most of what I have heard centered on other firms producing substantial components that previously would have been done by Boeing itself. Whenever there was an issue with some sub-assembly, the Boeing people felt that if the work hadn't been contracted out, the issue both wouldn't have happened, and would have been faster to resolve.
You sub contract out hardware assemblies (and software) when you start feeling market pressure and your cost has grown to be competitive.
Then you find, as a prime contractor, that you are still responsible for the product. You still have to maintain expertise. And, to keep the vendors in line you have to be able to at least threaten to bring the work back in house. All of this greatly reduces the cost advantage.
Now days, the concept is "only sub out work that is done better than you can do; where better includes technical performance".
What I think many people fail to recognize is that all aircraft vendors outsource quite a bit and have for a long time. Neither Boeing nor Airbus make the avionics. They buy them on spec from Honeywell or Smiths.
Back when I was a DS9 working in the ASL, we’d sometimes use a full software stack to try something out with pilots and airlines. But the result was a spec that went to Honeywell for implementation. And then when we got their output, we’d put those hardware boxes into our engineering sims and see if everything matched up.
I like the key attributes part. The engineers are paid by the hour, so they are incentivised to have the project running for as long as possible.
I don't think it's cheap to get an offshore team to do because the engineers might be cheap but the whole team + mgmt + reviewers are expensive. I have heard stories (from a person who did work in such environments) that the managers would charge a ton and do nothing basically. The work was handled by engineers and was reviewed by offshore people. Not sure what the managers were managing.
There are 2 advantages of an offshore team, especially India,
- Round the clock work
- Contractors can be hired for a few months and don't cost as much in terms of benefits that the company needs to provide to full time employees. So when you know work is going to pour in for 2 quarters, hire a bunch of a contractors. That way you don't have to pay for the equivalent on-site staff all year.
But obviously, this only shows up in quarterly financials and it's hard to judge the quality of the product over multiple years and iterations which is abused.
Not all execs care about "sustainable". They often have quarterly targets. To get some wins an exec could outsource, complete a bunch of tasks, then move to another company on a higher paid job, leaving former company to mop up.
See this a lot with sales also. Transactional client relationships promising the world to get the deal and hit commission multipliers. They then move to another company with "success stories" leaving implementation teams to reset client expectations.
It's worse than that. Lots of executive positions are bullshit jobs. They may or may not produce good results, and there may or may not be a way to determine that promptly. The important thing is to be able to fill up some slides in a deck with visionary plans. The engineers should have gotten them to discussing bike sheds.
I think it’s lack of long term thinking. The IT and HR departments at my company constantly come up with new shiny initiatives for saving cost. They centralize a lot of stuff and save some money that way. Over time it turns out that things don’t work but the initiator has already been promoted away and the next guy starts his own initiative. So there is this constant cycle of short term savings that cause long term cost. And it seems nobody cares about these long term effects because they all are focused on putting up a good short term show.
If you bought some stock, and were forced to keep it for atleast 5 years, the CEOs would have to behave a lot more carefully (=good for the company in the long term), than they currently do. With millisecond stock tradings, noone cares about long term future, because the stocks are sold and rebought and resold again and again many times by many trading houses and their financial products.
The whole stock market is just a speculation game now. With no real value defined behind it's price. All software and strategy is defined on finding stock at a higher price due to any reason. If you can't find a reason, make one. Like CEO's buying more stocks to limit the supply of their stock and propping up the "value".
It's all about the "sugar rush". Doesn't matter if the commodity is worth shit. The GameStop fiasco demonstrated that amply. It had no real viable plan, so it shouldn't have been worth a lot but you could artificially increase it's value.
They have to do it because VCs aren't going to stick around years for your organic growth. It was supposed to be a proof of viability and then proper quality product. With this model of money making, it's just the PoC.. no need for quality.
Cost cutting. Speed to market. Expectations of perpetual continuous improvements. Quarterly mindset. Inability to admit mistakes. Move fast break things.
It’s basically the same thing that led to the Volkswagen fiasco.
Unfortunately there’s so much code in these products, they start to believe they are SV style startups a fail whale situation has slim consequences when it’s completely opposite.
Sustainability doesn't matter when faced with a quarterly review. You do whatever to get it over the line in time for that and then don't have time to fix the snotball of hacks that is hidden.
I've certainly done this for sprints. I suspect if you examined code quality by when a ticket was started in a sprint, you would find it plummets a few days before the end.
As an American, I think that arguments about outsourcing that boil down to ability being all over the map don't feel right. Plenty of American engineers are all over the map.
The biggest difference is distance, the time zone split, and cultural differences.
I say this as someone that thinks limiting insourcing raises wages in the short term.
I agree, I've had plenty of colleagues whom really didn't excel in the field. I think for the most part with outsourcing as you say is the distance, but for me/us (team) it was cultural, we simply couldn't mix with the foreign part of the team. I've left the country out, because I don't want to point fingers, but it was simply about face, and no asking questions at all. To be honest, there were mistakes on both fronts, I feel like we as the engineering team in the country where the company was located was, felt entitled to our way of working, while the other party had to fit in.
wow, and here i am coding off and on since i was nine years old (port scanners, bots, blogs, ezines, chat servers, simple games, apps etc) cd and feeling like an imposter, so i don’t bother applying to tech jobs because i think i am not good enough and yet guys like the guy you worked with are getting all sorts of positions. maybe i should push myself more. having social anxiety sucks
Honestly, at least in my country it doesn't take a lot to get a programming job, at least if you can show you've got some skills. Though most companies won't take a look at you if you haven't got a diploma. If you've got the interest go for it. You cannot really lose anything by applying to some companies other than maybe a bit of time.
