Agile Offshore

Vincent Massol writes (and presents) about his experiences using agile methods with offshore development.

I read through the presentation and I think he's dead right. I forwarded it to my boss and my boss's boss. Here's hoping they read it. On the one hand, I hope against hope that we won't actually end up offshoring. But if we're forced to it, I hope that we're smart enough to actually follow the guidelines that Agile gives us.

Unfortunately, I’m not optimistic. Many of the recommendations here given are things that would actually increase the upfront cost of doing offshoring. And when visible, upfront costs are weighed against the hidden cost of failure, people too often choose the short-sighted route.

There’s an additional dimension that we can add on top of this: It’s an entirely different political situation depending on whether you employ developers offshore, or if you contract with a third-party development organization offshore (like WiPro, for example). If you have employees, then the company is responsible for making the whole thing work. If you’re working with a third-party, then the path of least resistance is to wait for the project to fail and then blame the contractor.

And if you’re hiring a third-party, how do you convince them to use your Agile process? In my (admittedly limited) experience, these outfits already have a development process that they want to use. They ask for a waterfall-style design spec and then expect to report back in six-months with a finished product. Are these ideas only useful if you’re hiring your own employees? And if you do that, aren’t you adding yet again to the cost?

Sigh. I really think the whole thing is 90% likely to be a miserable failure. And if you actually managed to do things correctly, I bet you don’t end up saving all that much money. Hiring good people (even offshore) costs more. Facilitating communication costs more. Doing it right costs more. I wish I had some real numbers to back that up, though.