Finding root cause of unsatisfactory Earned Value measurements

Describe how you would go about finding the root cause of unsatisfactory Earned Value measurements on a project, and what measures you might take to bring the project back into plan. Please be specific. Let’s say that two members of the team, working on unrelated tasks, have both run significantly over schedule in the same reporting period. One of them is working on a critical path task; the other is not. What do you look for? What questions do you ask? What might you do?

 I shall post my answer considering the fact that we are concerned with schedule variance only (not cost). First of all, I think EVM considers all activities as whole (to give complete project performance) and does not consider critical path activities separately. So the situation is little tricky. This calls for calculation of Schedule Performance Indicator specifically for the critical path and special consideration for the critical task. This doesn’t not mean that we need to forget the non-critical task – it has to be kept on track too. Bottom line – both tasks needs to be managed appropriately.

What do you look for? What do you ask?
Say, we are at 7th month of a 1 year project. First thing I would look for is to collect data from the Project Schedule to identify Earned Value (EV), Planned Value (PV) and schedule performance indicator (SPI) values for those two tasks separately. Questions asked would be –

1. How much work is actually done?
2. What was the planned value for the task for that reporting period?
3. How much work was supposed to be done?
4. What was the reason for negative schedule variance factor? Scope Change/Resource scarcity/Idleness/resource bottleneck or others?

What might you do?

Using the data arrived, I would draw separate EV charts that plots EV and PV values until month 7.

Managing Critical path task – The EV chart would give me the schedule variance. Say, it is 2 weeks for the critical task. I would look out if there was a total float/total slack allowed on the task. If yes, and if it is 2 weeks, then I would adjust the Project schedule accordingly. If there is none, I would look out for identifying total slack of any succeeding tasks and try to find if 2 week can be adjusted. If not, then I would have 3 options

1. Allocate more resources to the task and any succeeding tasks in the critical path so that the scheduled project finish date if met. (increase cost)
2. Negotiate with top management to reduce scope (reduce scope)
3. Negotiate with top project sponsors to allow for 2 weeks lag (increase time)
I would choose option 1 considering my commitment to the project and given that the sponsors didn’t accept options 1 or 2. Executing tasks in parallel is not a good option due the inherent risks that it might bring in toward schedule performance. Putting more money on resources will have to be justified as trade off to time and scope.

Managing Other task –
The only difference in managing this task would be the way I would look out for adjusting slack/float time. I might not really do such an adjustment, as there is no impact on the critical path. I would rather do research on answers to question 4 and try to identify the constraint that is delaying the task. I would then use the same analogy as with the 3 options (as above) to fix the negative schedule variance.