I would like to set IT organization wide control limits for a few basic project measurements:
1) Budget performance to forecast (% variance)
2) Effort performance to forecast (% variance)
3) First time quality, measured by % of test cases passed first time in system test (and user acceptance test)
Here is the problem, I don't have any history available, so instead of making a number up, are there "standard" limits that can be applied. I would imagine the limits might be different based on CMMi maturity . . . though we are just starting out, so we would be at Level 1.
Thanks. Saving Changes...
Sort By:
George JucanManaging Partner| Organizational Perfomance Enablers NetworkWoodbridge, Ontario, Canada
Instead of looking for “standard” control limits that, even if exist (personally I’m not aware of any with a reasonable wide acceptance), might have little relevance to your organization, I would suggest 2 alternate options:
1. Ask your executives what their target is, what is their tolerance to variation in budget, effort performance and first time quality level, and use those as your control limits. Alternatively, even if formal history data does not exist, people that have been with the company for a while can provide you with reasonable estimates.
2. Use a six sigma type approach, approximating the distribution interval with “best case” and “worst case” scenario (e.g. what is the max percentage under/over budget they ever delivered with). For a mid-maturity company a 2 sigma target seems realistic. Not exactly your case, but you might get an idea about what I’m talking about – defining an estimation interval and limits – from my article Estimation Blues on this site (http://www.gantthead.com/article.cfm?ID=227718)
Hope this helps. Saving Changes...
Michael WoodProject Manager / Business Analyst / Business Process Improvement Guru| Independent ContractorGig Harbor, Wa, United States
When all the dust settles management could care less what the project cost... if the prize is worth pursuing; usually many multiples of the cost of the project, then time and outcome are all that matter. Culturally they are also concerned with the process in that it is worth repeating. For this they look to the project participants for feedback... or lack of negative noise. So, your measurements should be whether or not the project achieved the outcomes it promised and that may not be known for a while. Assume for example that a project costing 1 million dollars will yeild bottom line benefits of 15 million... So what if it cost 2 million or 5 million to get done.. the real importants is that for everday the project is not done the company is losing some protion of the 15 million. Just something to think about. Saving Changes...
