Lessons Learned from Taking On a Project in Crisis
3 Apr 2010 edit
- Demand all technical source material up front, such as functional specs, mockups, work that's been done to date, etc. Give the client a fixed amount of time to deliver that source material, and don't make a decision about taking on the project until you've seen it. What the client can deliver in that fixed amount of time will shed a lot of light on the state of the project and whether their expectations are realistic.
- Set clear time expectations. Am I willing to work 16 hours a day? Am I expected to? Are there hours during which I'll be expected to be available? Am I willing to work on the weekend?
- Find out whether the client expects me to be available after the imminent deadline, and to what extent. The last thing I want is to snatch defeat from the jaws of victory by being unable to support the code I've written.
- Do not accept responsibility for commitments made on my behalf. The recruiter said I'd be available six hours a day when I told him four? Not my problem. The client committed to having a feature ready for review without consulting me? They probably won't make that mistake again.
- Ascertain the rest of the team's commitment to the project. If I'm expected to work long hours, will they be there during those long hours to get me what I need? Are there any constraints on their availability?
- Establish a single point of contact at the client, and make clear I'll be depending on them to get me any information I need and I'll be treating their decisions as final. Insist that they participate in all calls I'm expected to participate in.
- Limit the amount of work I do before I have access to all client systems I'll need access to: version control, testing environments, ticketing systems, etc.
- Insist on a ticketing system. I'm new to the project and I have a lot to get up to speed on. I don't want emails flying at me from all directions -- decisions and technical requirements need to be documented in a single place that everyone can see. This is my only insurance when someone wants to know why something isn't done, or why it wasn't done the way they expected.
- Insist on version control, even if it's something crappy like CVS. I'll need a way to make sure the rest of the team has access to my latest and greatest. FTP blows, especially when I'm FTPing to a server where another developer is constantly deploying a new build, overwriting my work.
What other advice do you have?