Sunday, November 20, 2005
Introduction to EVM - The Frosting
The example of last week presents us with some good and bad news. But how much more productive were we and how much behind schedule are we? The EV technique presents us with the results by calculating:
- Cost Variance (CV): CV=EV-AC;
- Schedule Variance (SV): SV=EV-PV;
So in our case we saved 440€ in work (CV) and we are 880€ behind schedule (SV; and since we are one day late this actually is correct since it represents the value of a day work of the two resources). If we calculate only the Accounting Variance (AV) we would report 1360€ (PV-AC). What is the meaning of this value regarding project performance? Would it be better if we where less productive therefore decreasing AV? If you know please e-mail me because I don’t.
Therefore, we can draw the following conclusions:
- CV greater than 0; This means EV greater then AC representing a better productivity than planned (if you prefer it can be interpreted as greater speed of work);
- CV=0; This means EV=AC representing the planned productivity;
- CV lesser then 0; This means EV lesser than AC representing a worst productivity than planned (if you prefer it can be interpreted as lesser speed of work);
- SV greater than 0; This means EV greater than PV representing more work done than initial planned;
- SV=0; This means EV=PV representing that same work done as initially planned;
- SV lesser than 0;This means EV lesser than PV representing less work done than initial planned.
The Earned Value technique calculates also two indexes:
- CPI (Cost Performance Index)=EV/AC; Represents the level of productivity or speed of the work;
- SPI (Schedule Performance Index)=EV/PV; Represents the speed of evolution of the schedule;
From the usage of these indexes, we can draw the following conclusion:
- CPI greater than 1; productivity is greater than expected;
- CPI=1; productivity is the same as expected;
- CPI lesser than 1; productivity is lesser than expected;
- SPI greater than 1; schedule speed is greater than expected (ahead of schedule);
- SPI=1; schedule speed is equal to what was expected (on schedule);
- SPI lesser than 1; schedule speed is lesser than expected (behind of schedule).
Next week: Introduction to EVM - The Toppings
J
Sunday, November 13, 2005
Introduction to EVM - The Cake
Never wonder why each week the project is always on budget but the project always finishes costing more or finishing latter?
Imagine that you have a three-day project with one task done by one resource. You plan accordingly and at day 1 you plan 8 hours, at day 2 you plan another 8 and at day 3 another 8.
On day 1 the resource reports 10 hours. That’s great you are ahead of schedule! The other day it reports another 10 hours. At this time you are ecstatic you have done the project and saved half a day (life could not be better). At day 3 the resource reports another 10 hours. Wow… What happened? You go talk to your team member and he tells you that the project was sub estimated and the work instead of taking 8 hours to finish each day it took 10. This is bad news. How could you have seen this coming? Always busy with so many projects and the accounting numbers never reflect what’s happening on the field.
This happens when analysing only the accounting variance (work planned minus work performed). However, there is a better way.
The Earned Value technique is broadly used as a performance measurement method for projects of any size. It enables a project manager to assess if there are variances from the base project plan that require some sort of corrective action.
This method presents an overview of the health of the schedule, cost and scope of a project and reflects the actual performance of the project. This is accomplished by calculating three main values for each task on the project plan:
1. Planned Value (PV): also known as Budgeted Cost of Work Scheduled (BCWS) is the value of the cost (calculated at the initial planned cost) of the work planned to have been done so far;
2. Actual Cost (AC): also known as Actual Cost of Work Performed (ACWP) is the value of the cost (calculated at the actual cost) of the work done so far;
3. Earned Value (EV): also known as Budgeted Cost of Work Performed (BCWP) is the value of the cost (calculated at the initial planned cost) of the work actually done so far. Some people like to think of it as the “physical progress of the work”.
The following figure illustrates the concepts based on our three days project:

Let us assume that our resource costs 100/h. The grey bar represents the initial plan. The key values are:
1. PV at day 1 is 100*8; at day 2 it’s equal to PV at day 1 plus 100*8; at day 3 it’s equal to PV at day 3 plus 100*8;
2. AC at day 1 is 100*10; at day 2 it’s equal to AC at day 1 plus 100*10; at day 3 it’s equal to AC at day 2 plus 100*10;
3. EV at day 1 is 100*8; at day 2 it’s equal at EV at day 1 plus 100*8; at day 3 it’s equal to EV at day 2 plus 100*8.
The figure bellow displays in a graph the three key EV values:

So lets see what would be our conclusions each day. Since AC is greater than EV this would tell us that our resource is doing more work to get the same results. At day 2 that situation maintains and at day 3 the same result. So you could have act since day 1 to try to solve the situation. Another curious conclusion is that if we look closer at the results EV and PV are always the same. This tell us that we are actually on schedule and the project is going to finish at the planned date since the work actually done is always the same as the work planned. Great, now we can get a clear picture of the current project status instead of just hopping for the best.
Lets look at one more example.

Resource 1 costs 100€/h while Resource 2 costs 10€/h. As we can see, we are currently one day late and the current date is Day 3. Instead of starting working at Day 1 the resources only started at Day 2 and reported only 4 hours each instead of 8. The figure bellow displays in a graph the three key EV values:

So PV at day 1 is equal 100*8+10*8. PV at day 2 is equal to PV at day 1 plus 100*8+10*8. EV for day 1 is zero since no work was done. At day 2 EV is EV at day 1 plus 100*8+10*8. AC at day 1 is zero, at day 2 is AC at day 1 plus 100*4+10*4.
These values represent two things:
1. Since the work actually done (EV) is greater at day 2 than the work reported (AC) then we are more productive than expected;Since the work planned (PV) at day 2 is greater than the work reported as finished (EV) then we are behind schedule.
2. Since the work planned (PV) at day 2 is greater than the work reported as finished (EV) then we are behind schedule.
Next week article: Introduction to EVM - The Frosting.
J
Friday, November 04, 2005
Earned Value Management Applied to Small Projects Case Study

The project plan consisted of a total of 34 tasks and 4 people composed the team.
Each Friday the timesheets were submitted so the project progress could be assessed using EVM at Monday morning. The indexes were analyzed at the phase level and predictions were generated.
For illustration only purposes we present bellow the progress of the EVM indexes at the top project level and the main responses.

The actual analysis was done at the phase level but this figure is a good example on how to act based on the CPI and SPI thresholds.
In this project the most important variable to control was cost and the project predictions were mainly cost estimates. The estimates were calculated at the phase level and added to get the total cost estimate. The evolution of the cost estimates are depicted in the graph bellow.

From the graph I would like to emphasize the following facts:
- After 50% of the work the predictions were within a 10% interval;
- The highlighted section of the graph depicts a zone that predicted a big cost overrun if we didn't do anything. Due to the actions undertaken by close risk monitoring and control the final cost was well on target.

The graph shows that 80% of the predictions had and error inferior to 10% when compared with the final cost.
The usage of EVM in small projects is most valuable since, was we could see in our case study, the oscillations in cost and schedule in small projects can happen very quickly, demanding very close monitoring and control.
EVM provides powerful early warnings that used in conjunction with a good risk management process enables project managers to detect deviations from the targets and act accordingly.
Based on the concepts of EVM the project manager can calculate predictions of cost and schedule that can be very accurate.
EVM can be applied to small projects by using methods that with little effort can provide the full power of EVM. These methods are:
- Top-level EVM analysis done at the project life-cycle phases;50/50 rule or similar for progress report;
- Using index thresholds triggers linked with a risk management process;
- Use adjusted SPI instead of SPI.




