6 Reasons Your Favorable Performance May Be Too Good To Be True

You know the scenario well: it’s the end of the period and now you eagerly await the results of the latest schedule status.  Will the status align with the planned and actual costs?  You process the data and the results spit out of your EVM software.  Ta-da!  Your SPI or CPI are stellar!  Congrats!  But wait…a savvy PM knows when favorable variances and trends are deceiving.  The issue is much more complicated than just overestimating the percent complete progress for a few activities.  It emphasizes importance of true data analysis and non-complacency in relying on the numbers all the time.  Do you know how to spot deceivingly favorable performance?

While positive performance is generally a good thing, there are times it can come back to bite you…eventually.  Let’s take a look at the lesser known causes of deceiving performance numbers so we can determine how to best avoid them:

1. Poor Scheduling.  I doubt this one is much of a surprise, but it confounds me how many projects have schedules so poorly developed and thought through.  Activities out of sequence, as well as shoddy durations, man-hours, or materials estimates can all cause positive variances that are false in any given period, and may continue for several.
2. Devious Scheduling. This is different from poor scheduling in that it is done on purpose.  Specifically, this is done by baselining activities too conservatively (too long, too many hours, too far into the future), and doesn’t come close enough to reality.  Perhaps the PM did this on purpose to look like a hero by coming in well ahead of schedule or under budget.  The danger is that it also hinders the ability to really measure progress.  
3. CPI/SPI Out of Sync.  Sure, maybe you worked fewer hours overall, but you also didn’t get some of the work done.  How can this cause false positive performance, when you can’t earn value on work not done, you ask?  One answer is Level of Effort activities.  LOE will earn value whether staff is working or not.  Another answer is skeleton crews.  An example would be when excavation is still underway (using a crew of 3, instead of 5), and the 4th and 5th members are on vacation—along with most of the office staff (who are budgeted as LOE) for the holidays.  The forecast may show a favorable trend with a decline in total cost, but this disappears when everyone comes back to work in January and then puts in overtime to make up the work that wasn’t done over the holidays!
4. Wrong Tasks Progressing.  This is where Critical Path analysis (CPM) is important, as it reviews the mix of performance in the Control Accounts.  If activities are being work out of sequence, it might look favorable for now, especially if the working activities have a higher budget than the original planned activities for this time period.  However, it could result in added time/cost later.  This is your classic case of the cart before the horse.  A common example is developing too much training material before the final product is done, which then causes rework.  
5. Crashing the Schedule.  It doesn’t have to be a full-on crash, in fact it could be just a few activities with larger budget on them.  The SPI will show the project is ahead of schedule (either current period or cumulatively), because you threw more resources at these activities, although now the EAC is sky-high.  But hey, you’re gittin’ it done!  Here’s a tip: you should ALWAYS review your CPI and SPI in relation to each other.  
6. EV Method Misalignment Between the Baseline and Active Plan. Most commonly this crops up in the form of a negative variance, but it can be found to cause positive variances too.  For example, the baseline is level loaded (linearly) for a 4 period task, but the scheduler mistakenly statuses it as 50/50.  The 1st period will have huge positive variances since you earned much more than planned.  But then it will be followed by 2 periods of negative variances, until it catches up in the 4th period—where reports will finally show whether you were truly efficient or not.  

Ultimately, deceivingly positive performance boils down to either false information, or an anomaly that cannot be sustained.  Experienced, knowledgeable Project Managers know these are some of the reasons you cannot work in a vacuum with EVM.  Performance must be assessed in relation to individual tasks, groups of tasks, the entire schedule, the baseline time-phasing, the EAC, as well as the current period and cumulative to date information.  It’s a symphony of data—a system, and can only be effectively managed when treated as such.  Incidentally, this is one reason your EVM software is so important—a good package will offer many ways to easily slice the data so you can see these issues, because sometimes you have to dig pretty deep to find them.

Obviously not all positive performance is caused by these things!  Sure enough, your project MAY be excelling.  But you’d better know the difference.

Do you have any other examples of false favorable conditions?  Please feel free to leave a comment, and check out the next blog, “How to Truly Measure Work Accomplished”.

- Melissa Duncan (About Melissa)

Editors Note: We generally post twice a month, for December we will be posting only once and return to twice a month in January 2015.   

Read the previous posting - The Alternative to Earned Value Management: it’s called being FIRED!

Subscribe to Earned Value Exposed

Earned Value legend as there may be a few of us that don’t yet have this memorized…
CA=Control Accounts
CAM=Control Account Manager
CPI=Cost Performance Index
EAC=Estimate At Completion  

EVM=Earned Value Management

EVMS=Earned Value Management System
EAC=Estimate At Completion 

IPMR=Integrated Program Management Report
LOE=Level of Effort
OBS=Organizational Breakdown Structure
OTB=Over Target Baseline
PC=Project Controls
PM=Project Manager
PMB=Performance Measurement Baseline
RAM=Responsibility Assignment Matrix
SPI=Schedule Performance Index
WBS=Work Breakdown Structure