Adam Creighton, Computer and Video Gaming: The reality of outsourcing ...

Other Blogs:
American McGee
Beautiful Pixels
Hideo Kojima
Hitachi Wasabe's Xbox 360

Xbox 360 Marketplace Updates: (MajorNelson.com)

Adam Creighton, Computer and Video Gaming (Subscribe)

Tuesday, May 08, 2007

The reality of outsourcing ...

Outsourcing.

Take a moment and evaluate your response to that word.

What does it mean? In practicality, how does it play out in your vertical market, and how does it impact you?

No disrespect meant, but the video game industry thinks it's very special. And I think it's relatively late to the party and behind parallel efforts from Hollywood and Corporate IT.

But it thinks it's doing things for the first time. (To be fair, any vertical market thinks they're special.)

Like with outsourcing. There are a lot of video game companies just now beginning to explore outsourcing, what that means, and how it benefits the company, the team, or the game.

There's some misunderstanding about what outsourcing is and isn't, its benefits, and what the video game industry is doing recently with it.

First, to level set, realize outsourcing does not mean, "Losing jobs to India."

Outsourcing, at its core, is "sourcing outside". That can mean outside of the local group, but still within the local office; or outside of the geographic office, but still in the company; or outside the company, from another company that may or may not be local, regional, or even domestic.

The "losing jobs to India" is a perception, and only a facet of outsourcing. To be fair, some people have suffered that actual pain of that outsourcing. And then you have things like Bank of America laying off staff and making their severance dependent on training their overseas replacements. That PR doesn't help.

But I've managed outsourcing. At its core, outsourcing is figuring out how to reduce cost and still get acceptable quality.

So, for example, when I inherited a project that realized "cost savings" by off shoring day-to-day support, but that support wasn't quality? I had the vendor replace the support. Twice. Then, since they failed that, I moved to an on shore model, requiring the the same off shore rate to the vendor, because they were the ones who were not able to provide adequately trained or managed off shore resources.

Or when I worked for IBM somewhere when IBM was moving away from charging internal "Blue Dollars" for internal work, I would beat the bushes for resources who had extra cycles and relevant expertise, and have my management negotiate with their management about providing stuff we needed, and vice versa.

I've also manage the "RF[n]" process, ad nauseum -- Request for Proposal / Information / Quote -- to find adequate vendors for plug-in, one-time, or special use pieces that don't make sense for me to redirect my folks ("don't make sense", for me as a development manager, includes the effort wouldn't give my folks skills they can use in their careers, regardless of whether it's for me or the company).

So, what makes for good outsourcing candidates?

It's not simple, but good candidates include projects that can be broken into discreet modules or chunks that are not real-time dependent. So, libraries often make good candidates. Assets (art, 3D, animation, and others that have hook able templates) make good candidates. The candidacy pool for outsourcing efforts grows as you move from international to domestic, national to local, and local to in-house, different group. Inflexibility of a development methodology (RUP, Agile / Scrum, etc.) can severely reduce the pool of candidate items.

What makes for bad outsourcing candidates?

Again, not simple, but items that require real-time resolution need to be evaluated in terms of the Service Level Agreement (SLA) for acceptability nationally, and especially internationally. (Support a business application that requires next-day resolution for any issues? That's a high-risk candidate for overseas outsourcing.)

And some video companies get this, and are already practicing it.

In 2004, I attended an excellent session at that year's Austin Game Developers Conference, where rockstars The Animation Farm (still, animation, 3D, storyboard, mocap assets) and Critical Mass Interactive (turnkey development and management solutions) talked about what they were doing -- and what worked -- in the outsourcing arena (frankly, much of which involved overseas outsourcing and management).

And now, big companies are making concerted recent outsource management efforts.

Blizzard has announced the formation of an Austin customer support office, roughly 500 strong. This is fairly intelligent outsourcing, because it takes advantage of Austin's rich IT support infrastructure (and job pool), and may well react to California legislation reclassifying several IT roles as hourly, which can get very expensive for a corporation. And it creates some subset of 500 jobs in Austin (I have to think a portion of them are going to be willing or encouraged relocations from Irvine, CA, and I'm angling for a management job, so you have to remove that from the mix).

And then, a bit under the radar, in February "Midway Announces Creation of the Central Outsourcing Group (COG) to Streamline Outsourcing Efforts" (is Epic ticked?), which is geared toward ""consolidating all product development outsourcing initiatives into a centralized group" -- and managed (coincidentally?) out of Austin.

And off shoring of work is happening in the video game industry. You have a company like French-headquartered Ubisoft create Ubisoft Shanghai, responsible not just for pieces, but entire Ubisoft titles. And LucasFilms is rumored to have a massive LucasFilm Shanghai group working on a secret 3D Star Wars entertainment project. (When I worked with Peoples Bank of China, they were adamant their country's greatest natural resource "is our people", and were far more aggressive than any group with whom I've worked on parallelizing development efforts.)

All this is to say (in a rather rambling manner) that outsourcing is complex, has benefits, and the gaming industry could probably learn some lessons from their non-gaming IT counterparts who have been struggling with it for years (and should give themselves some grace because their counterparts are still struggling with it).

Labels: development process, rants