The Hidden Costs of Offshore Outsourcing

CIO Magazine ^ | Sep. 1, 2003 Issue | STEPHANIE OVERBY

[Dominic Harr]

And so it begins. Finally the press begins to look into the results of the projects which have been off-shored. We're not able to find a single success story.

The 'offshoring' fiasco will soon take it's place beside the Dot Com and Y2K panics.

These pointy-haired bosses in American IT corps are really, really something.

[dirtboy]

They left out two other critical costs - the costs of alienating your remaining American employees, and the possible costs of a backlash against the company as consumers get more and more irate against outsourcing. And those may end up being the greatest costs of all.

[Dominic Harr]

Forgive me anyone I'm forgetting to ping.

But I know this will be interesting to ya'll.

[Dominic Harr]

Actually, the do brush all that in the section, "The Cost of Layoffs".

But you're right, the 'good will' loss is indeed real.

[Lazamataz]

I'm actually getting the impression that this article is showing one or two successful offshore projects, but they have to pay nearly the same money to get it working.

So why not keep it in-nation?

[stainlessbanner]

Perhaps the greatest cost of off-shoring is erosion of the existing customer base during and after the transition.

[snarkpup]

"A good American programmer will push back and say, What you're asking for doesn't make sense, you idiot," Zupnick says. "Indian programmers have been known to say, This doesn't make sense, but this is the way the client wants it."

Based on personal experience, this is one of the strongest points of the article.

[Dominic Harr]

So why not keep it in-nation?

Exactly.

And then, there's that really evil little problem -- time to market.

If it takes us 2 years, and it takes them 4, in the computer market, that's a massive failure, to get beat to market by years.

Notice MS has pushed the release of their next OS release, Longhorn, back another year? And they use *all those indians*.

Good. Linux can use the time.

[Dominic Harr]

Perhaps the greatest cost of off-shoring is erosion of the existing customer base during and after the transition.

Bingo.

It's actually become a selling point for many contractors here that they use only American resources.

That's becoming a big selling point.

[Lazamataz]

Notice MS has pushed the release of their next OS release, Longhorn, back another year? And they use *all those indians*.

Well, to be fair, I think MS has ALWAYS been years behind it's projected release date.

[Snuffington]

Most of the Indian offshore companies are ISO certified and at Capability Maturity Model (CMM) Level 3 or 5. "If your own staff can't get used to working at that level, you're going to have a major disconnect," Raspallo says.

Having worked at a company implementing CMM purely to be able to offshore, I have to notice that this article is missing some major productivity and opportunity cost related to this.

CMM tends to create a frozen-in-time approach to software development (No, it doesn't have to, but that's the usual result). This stifles innovation and process improvement (Yes, I know CMM 5 is all about a self-improving process. I also know that its primary purpose is to convince share holders, rather than deliver solid results.). It also tends to have a result on some of the most innovative and productive developers akin to the result kryptonite has on Superman - it either kills them or drives them away.

[brownsfan]

And why is it none of these geniuses can see the "intangibles"? I can see 2 really big ones without trying hard. 1 - Lack of personal investment from employees. If you worked for a company through a brokered arrangement, might you lose the personal stake in the company? No loyalty?
2 - If you send all your jobs to India or China so you can pay $1.25 an hour, and other businesses do too, who will buy your goods?! If all the jobs are overseas, and they all pay poorly, won't you eventually have no one to buy what you sell?
Clearly, most of these guys want to make a profit today, future be damned.

[Starwind]

"Indian programmers have been known to say, This doesn't make sense, but this is the way the client wants it." Thus, work takes more time and money to complete. And a project that's common sense for a U.S. worker?like creating an automation system for consumer credit cards?may be a foreign concept offshore. Additionally, offshore vendors often lack developer experience (the average experience of offshore developers is six years).

It's more than lack of developer experience with a given technology, it's a lack of experience at solving real-world problems.

They often to how to code, but seldom know what to code.

They have little experience anticipating what can go wrong and how to build in testability, reliability, maintainability, extensibility, and scalability, portability, (and a few more 'ilities). There is a lot of 'talk' about how OO or .NET, J2EE, etc automates this. But it's 'talk' and what is needed is experience in product and technology selection as well as system architecture. There is a huge dfference between building one-off apps for some end-user and a product-like system for 10,000 of users/customners. Experience even the smartest Indian developers won't learn in Hyderabaad for many years yet.

[Taylor42]

I've seen our management team play with productivity numbers to make our "Best Shore" projects look successful.

[Dominic Harr]

I believe 'process', specifically CMM and the like, is the next big BS Hype cucle, after this one finally plays out.

'Repeatable processes', even if those processes are crap.

There's only one process that works -- get good people, give them the tools they need to do their job, then punishing them when they fail to meet deadlines and requirements, or rewarding them when they succeed.

The problem is the 'reward' part. These managers and executives only want to pay themselves. They want to give themselves big bonuses to management, and pay the actual skilled workers $5,000 a year`.

[Dominic Harr]

Experience even the smartest Indian developers won't learn in Hyderabaad for many years yet.

Indeed.

One way of thinking about it -- it takes, I'd say, a minimum of 20 years to really train a high-quality developer.

Most of us start at around age 12 to 13, then spend at least 10 years learning about PCs, all the sublteties and nuances, how the darned machine really works, what a useful, good screen looks like, what kinds of things can go wrong, etc. Writing simple code, random number generators for games, graphics editing, doing beta tests for games and other software (a *brilliant* bit of experience for any developer), and on and on.

Then after about a decade of that type experience, we begin serious programming. And then it takes what, minimum 5 years before we're really what I'd call 'good' at basic, simple development.

To become an 'ace' takes at least another 5 years. And also the ability to learn multiple languages, and adapt well to all the new technologies that will come along.

They're taking kids straight out of college with minimal computer exposure and waving their magic wands, viola!

We have a developer!

Um, no.

[CodeMonkey]

How about the biggest cost, if you outsource all of the good jobs to countries where the same job pays 1/3 at best then who can afford your products? The reason that Ford paid "obscenely high" wages to his line workers was to get capital out there to buy his products. $5 a day back then was considered an unheard of rate. Guess what, that was one of the only things that Ford was ever right about.

I suppose in the end where most hardline capitalists differ from liberal capitalists (ie Locke, Smith, etc) is that the liberals believed that there had to be limits on what the market can do otherwise it may end up ripping society apart. Outsourcing taken to its natural conclusion would cause a socialist revolution in the US as millions get sick of watching Indians and other foreigners getting our jobs.

[Southack]

"And so it begins. Finally the press begins to look into the results of the projects which have been off-shored. We're not able to find a single success story."

I didn't see any real stories of project failures in this article. Instead, what I saw were some of the financial realities of offshoring being exposed.

What you are speaking of would make for a much better article.

Let's talk about the Chinese theft of the optical router source code from Lucent's offshore outsourcing failure.

Let's talk about specific, large projects that have been sent offshore only to fail, be cancelled, or wind up costing more than they would have been expensed if they had remained in the U.S.

[Dominic Harr]

Outsourcing taken to its natural conclusion would cause a socialist revolution in the US as millions get sick of watching Indians and other foreigners getting our jobs.

I appreciate what you mean, but I just don't like that thinking.

If the Indians *can* do our jobs better, cheaper, then they should.

We should change careers, or improve ourselves.

Now, I do *NOT* think that is happening here. I think the only reason the Indians are temporarily landing these contracts is because of wide-spread corporate corruption and stupidity. Literally, these executives are giving themselves million-dollar bonuses with the money they save by paying their employees $5k a year. Sure, the company may be failing, but the executive gets rich, quick.

Just remember, the "Make Millions Moving IT Jobs To India" hype came to you from the same crowd that was screaming, "Make Millions Investing in Dot Coms" and "The Y2K bug will have planes falling out of the skies".

The Dilbert, "pointy-haired boss" is a true American Icon now. They're a public joke, and their idiotic decisions are such a joke they're the basis of one of the best cartoons (Dilbert) and some good movies (Office Space) and a great brit TV Show (have you seen "The Office"?).

[WilliamofCarmichael]

Another productivity killer is high turnover at offshore vendors. Attrition rates climb as high as 35 percent in India, according to the National Association of Software and Service Companies. "Unless you can somehow address that in your contract, you're paying for someone to learn your product and then they're gone," Zupnick says. Turnover can cost an additional 1 percent to 2 percent.

There are so many true things but this is the one I was looking for. With the expected flood of IT going off shore this is not going away. So India forces employees to stay on the job? Would the "world's biggest democracy" do that?

Turning over poorly documented work in progress can cost a lot more than an additional 1 percent to 2 percent. I am talking about long-term projects involving teams of developers.