Tuesday, June 10, 2008

In light of today's earthquake in China...

A month ago on LinkedIn, Tatyana Kanzaveli, Senior Sales and Business Development Executive asks:
In light of today's earthquake in China... wouldn't this trigger your thoughts on multi-sourcing/disaster-proving your outsourcing strategies?
Hope to hear more insights on this topic at upcoming International Software Development Outsourcing Conference
My answer:
The earthquake in China and the cyclone in Myanmar are all examples of the potential danger of single sourcing. Of course, given the immense benefits of outsourcing, I would not argue to remove outsourcing all together. In general, I think the benefits of disaster preparation are woefully under appreciated. Here are a few of the things I would keep in mind before outsourcing:

1. Are you outsourcing your competitive advantage? I would be wary of outsourcing anything your company does best. The best candidates for outsourcing are non-essential functions that someone else is clearly better at.

2. If a site were to go down, how quickly could you bring up production at another location? How critical is it really? In general, customers are very forgiving when you say that an earthquake leveled your location in China, but this is a chance to wow them if you can get back to 100% quicker than anyone anticipated.

3. What is your data retention strategy? Do you back up all data onto a server in the US on a nightly basis? If a foreign government takes over your business, do you have the capability to wipe data off remote drives? Do you have a remote detonation device? (Just kidding).

4. Economies of scale: If they are very small, it is much more cost effective to multi source. If they are great, it would lead more highly to single sourcing.

5. Too much power. In general, with one supplier you give up lots of power to just one party. Having two parties will allow you to compare them against each other to see who provides better value. It’s very hard to compare a single supplier with others because it is basically guesswork.

0 comments: