CTOing - Evaluating Contractors
I'm re-posting this from my blog. Don't worry, I'm not going to make a habit of recycling things here, but I think a few posts are a good fit so I'll move those over here between other posts.
A question was recently posed to a CTO mailing list I'm a part of as to evaluating contractors. The poster was looking at frontend web development contractors and trying to determine the best way to evaluate them. In case you find yourself in a similar situation, here is my (slightly edited) response:
OP,
I’ve done several types of evaluations. I've found that the two below seem to work best for me.
I greatly prefer using real world issues. I usually setup a small agreement (4-8 hour contract) with appropriate NDAs. It ensures the contractor receives value for their time and gives me a good feel for how their work will be. I'll then have the contractor work a real issue/ticket we have open on the system they'll be working with. Just make sure to include time for them to get setup with their local development environment along with appropriate instructions.
If that’s not possible, I’d recommend having them do a sample project thats similar to your core stack. Although exact frameworks and libraries wouldn’t be a concern for me. I'll usually have someone create something simple, a guitar tab site is a fun favorite of mine, and dictate one or two tools. Then I’ll let them decide the rest. I find that seeing how someone architects a project with a lot of decisions left up to them tells you much more than having them follow an exact spec.