This is an entry that extends the post on microprojects by two additional prioritizing strategies.
Strategy: Cover your ass
This is the strategy suggested by the previous post. After preparing a list of milestones and their estimates, you pick the next most problematic milestone and work on it. A list of tasks ordered by this strategy helps you to “fail fast”: in less than half of estimated time you will know, whether you will succeed or bust the budget – even when little is known about the concrete implementation or the esimates are off by some amount.
Strategy: Most value first
In lot of projects, this is the strategy used by customers. All features not absolutely necessary to achieve the goal are cut or declared optional. If you look at minesweeper: you can play it without the highscore, the timer, the modifiable field side or even probably without the random component (i.e. make 99 fields), but not without the mines. After you determined that your budget is too small, you know what the customer can live without and if you have the option to cut features, then this is probably the strategy for you.
Strategy: Most painful, when omitted
This is the strategy best applied before the pain is real. In contrast to other two strategies, it does contain hard to quantify criteria like:
The cost to implement them is non-linear and not directly visible. The temptation is big to use time and money to create more profitable features instead. They can be prioritized by:
- probability of occurence
- damage in the case of occurence
- implementation cost
- growth of the above factors with time
This is a lot of work for a single task – most likely you will setup project wide guidelines and default scenarios that will be reviewed by recurring audits.