Every projects most serious enemy is the feature creep. Somebody - with ample political support or aggressiveness - pushes his favorite function into the scope - not only delaying the project and increasing investment, but often also complicating the process and system operation which is a much greater sin because it will punish the buyer of the software permanently. So how to avoid that?
It seems that no method is really useful, it's a complex issue. What can you do?
Somewhat you have to prioritize; even with the largest budget, best intention and ninja power you can't do everything you are asked for. Priority means that you put the more important things forward and the less important at the end of the list. For everybody the most important is her stuff. So you have to define importance too. A thing is more important than an other if its cost benefit ratio is higher. Cost is relative(!) easy to define but with benefit you are stuck again. Benefit is - normally - defined by some upper authority and it'd clear and measurable. The most obvious measure of benefit is money saved or money earned but it also can be customer satisfaction or processing time... The more easy the measurement is the easier to compare things and evaluate results. You have to put a considerable effort to define what what you measure, because the objectives are not always clear. E.g. "improve customer satisfaction by 30%" is a good target but it doesn't say anything about how we measure customer satisfaction. So there is a twofold agenda before starting any project; you have to clarify the objectives and set the KPIs. If you have it cleared and approved then you can evaluate every function based on the KPI. Does it contribute to the to the objective? Can't be clearly measured by the KPIs. And doesn't it improve the KPI? Does it create value? End if the answer to these questions is yes then how much is the effect of it? By doing so you have better chance to make discussions professional and factual.
A good tool to the analysis of objectives is the CTQ flowdown diagram.
No comments:
Post a Comment