Weathering the Storm of Cost Variance Analysis: Part One

My last post talked about the Bermuda Triangle of project management, variance analysis.  Specifically, for Earned Value Management reporting, we discussed schedule variances.  In my opinion, schedule variances are the tougher of the two variance types to analyze.  This week it’s time to complete your training and show you how to navigate through the write-up of a solid cost variance analysis.  In this situation, your actual cost is out of line with your performance numbers.  It’s either costing you more or less to produce that widget and everyone wants to know why.  What they really mean is “what are you doing to screw this up?”  Cost variance analysis can be tricky.  In a negative variance, you’re trying to communicate the reasons for performance numbers that make it seem like your ship is sinking, yet still calm the client’s nerves, and make it clear to management you know how to fix the hole in the boat.  In a positive variance, you’ve got to convince everyone that this trend is still temporary and to not get carried away with taking money from your project.  Welcome to the eye of the storm!

When reporting these kinds of Earned Value metrics, how does one satisfy management, the client, and the project in a cost variance write-up?  As we discussed with schedule variances, the formula is the same: keep it simple.  Make a list of the tasks that are causing the cost variance and address them in a factual sense.  Perhaps labor was more costly than originally planned?  Do not get into the common mistake of saying “task took longer than planned” because is incorrect when dealing with a cost variance!  That phrase is describing is a schedule variance.  What you mean to say is, “task required more effort than planned”.  Quantify the words “more effort” by defining it as more man-hours or overtime, etc.  Be specific about what took more effort, why, and what corrective action needs to be taken.  Once again, be wary of talking about schedule, and don’t forget to only discuss variances associated with work that was done and its associated cost. You can’t have a variance on work that isn’t in the baseline.

If you’re not true to your variance write-up, you may not get the “help” you need to course correct the project.  So be honest about what is occurring to cause the variance and share your plans for turning it around. And if there is a continuous negative trend on the project, it’s OK to admit it and say that it will be reflected in the project’s EAC.  I’ve seen quite a few PM’s hanged by their shoestrings when trying to “hide” the real reasons for variances.  Be politically correct, be eloquent even, but don’t lie.  They’ll catch you eventually. 

Remember, in order to write a proper cost variance analysis; you have to start with the correct cause for the variance.  To help get your feet wet and start you in the right direction, let’s take a look at some common causes for cost variances:  

• Rate changes (i.e., labor, overhead)
• Vendor discounts or price increases 
• Quantity discounts
• Material cost changes
• Requirement changes
• Inefficiencies in labor (i.e. extra hours needed to complete what was planned, this can be from inexperienced craft available from the labor union…etc.)

Click here to view: Part 2 of this blog article, where look at some examples of these issues

Our next posts is Scale your EVMS and obtain the Biggest Bang for Your Buck

- Melissa Duncan (About Melissa)

Read the previous posting Earned Value Exposed: How to Avoid Being Sucked into the Votex of Schedule Variance Analysis. 

Subscribe to Blog  

Earned Value legend as there may be a few of us that don’t yet have this memorized…

PM=Project Manager

PC=Project Controls

CAM=Control Account Manager
CPI=Cost Performance Index

EVM=Earned Value Management

EVMS=Earned Value Management System

EAC=Estimate At Completion 
WBS=Work Breakdown Structure

OBS=Organizational Breakdown Structure
SPI=Schedule Performance Index
IPMR=Integrated Program Management Report

Weathering the Storm of Cost Variance Analysis: Part Two

We are discussing the complexities of Earned Value Management related to cost variance analysis and have covered the basics in Part 1.   Let’s take a look at some real examples from real projects.

Good CV – Example #1 (Overrun situation)
Issue: The cost of preparing the xx document increased due to overtime required to process the high number of comments from the customer and the Board of Directors from the review process. 
Status: The additional cost will not be recovered.  The overtime allowed the project to recover the slip in the schedule.  Overrun has been trended and has been reflected in the project EAC.
Comments: 

• The issue explanation is clear and easy to understand, yet brief.  It states what and why.
• The status is clear and direct.  Two additional statements were added to show the related impact.

Good CV – Example #2 (Underrun situation)
Issue: The favorable cost variance is due to savings that resulted from consolidation of two procurement documents.
Status:  Cost savings have been documented and trended in the project EAC.  However, increased costs are anticipated in the fourth quarter due to projected increases in engineering and subcontract execution costs.
Comments: 

• The issue identifies the scope and why the variance occurred.
• Statusing that the trend has been documented is appropriate.  Since there is a future impact that will offset this trend, the additional statement is necessary.

Poor CV – Example #3 (Underrun situation)
Issue: The baseline reflects level-of-effort evenly for the entire FY.  The current schedule does not reflect a level-of-effort forecast.
Status:  Work will be performed in the last half of the year, decreasing the underrun.
Comments: 

• The issue statement does not describe what specific baseline scope is underrunning or why.  “Level-of-effort’ is also inconsistent with tracking progress by performance and deliverables, and should not be used.
• When the work is performed will influence the SV, not the CV, so the status statement is incorrect.  Likewise, if the work hasn’t been done yet (since it will be done later), there can be no underrun since there was no performance, and performance should not be reported on the schedule.

Poor CV – Example #4 (Underrun situation)
Issue: Savings from efficiencies in performing the valve scope.
Status:  N/A
Comments: 

• The “efficiencies” referred to should be defined.  If the word “efficiencies” is used, describe specifically what was done better/cheaper.  Ask yourself if this “efficiency” can be applied at other locations.  Usually, the best response is to simply identify the task that the savings occurred on and state why or how it caused the underrun (such as replacing fewer valves than planned).
• Since this is an underrun, no corrective action is necessary for ‘status’, but do state that the underrun has been/will be reflected in the EAC or that it will be offset by higher costs later.

Poor CV – Example #5 (Underrun situation)
Issue: Underrun from cancelling turbine work.
Status:  N/A
Comments: 

• The issue statement needs to be more specific and better defined.  Cancelled work cannot have a cost variance (must have performance to have savings) so describe the trend instead.  What happened and why is needed in the explanation (such as savings from consolidating two requirements into one rather than cancelling one of the requirements).
• Since this is an underrun, no corrective action is necessary for ‘status’, but do state that the underrun has been/will be reflected in the EAC or that it will be offset by higher costs later.

Poor CV – Example #6 (Overrun situation)
Issue: Overrun due to accrual error.
Status:  Error will be corrected next month.
Comments: 

• Even though it may be a valid reason, the “accrual error” variance explanation is used far too much and can over-shadow problem situations.  Many times accrual errors are not taken seriously enough, and can cause FY funding problems as well as a variance.  If the accrual is large, state what work scope was impacted and why the accrual was in error (such as entered twice or that it was entered incorrectly).
• Correcting the error next month is an appropriate statement if the accrual error will be reversed dollar for dollar.

Poor CV – Example #7 (Overrun situation)
Issue: Cost of the work was more than planned.
Status:  Overrun has been entered into EAC.
Comments: 

• The issue needs to be more specific.  Identify the specific work and state what was in the estimate or the assumption used vs. the actual result.
• Status “entered into EAC” is OK if it cannot be made up.

Keep in mind that cost variances are not impacted by changes to the EAC, remaining budget, funding, or future work scheduling.  This is your opportunity to communicate the activities causing the variance, the issue, the impact, and corrective action if necessary.  Focus on those simple items and you’ll get through the process each period relatively unscathed.  

This completes our two part series on variance analysis.  Do you have any great examples of what to write or not write?  Please share in the comments below.  Our next post's title is Scale your EVMS and obtain the Biggest Bang for Your Buck.  Here we will look as how to scale an earned value system.  

- Melissa Duncan (About Melissa)

Read the previous posting Earned Value Exposed: How to Avoid Being Sucked into the Votex of Schedule Variance Analysis. 

Subscribe to Blog 

Earned Value legend as there may be a few of us that don’t yet have this memorized…

PM=Project Manager

PC=Project Controls

CAM=Control Account Manager
CPI=Cost Performance Index

EVM=Earned Value Management

EVMS=Earned Value Management System

EAC=Estimate At Completion 
WBS=Work Breakdown Structure

OBS=Organizational Breakdown Structure
SPI=Schedule Performance Index
IPMR=Integrated Program Management Report

How to avoid being sucked into the vortex of Schedule Variance Analysis: Part One

Welcome to the Bermuda Triangle of project management, where even the best PM’s can get lost and fired along the way.  Writing a schedule variance analysis for a certified EVMS can be painful, much like a love triangle.  You must please the client, your management, and the project—all while trying to keep the political fire under control.  The project is in hot water and you have to ease the minds of your management and the client, but give it enough teeth in order to get the assistance you may need to get the project back on track, such as additional resources.  The client wants ALL the details, while management does not want you to share any of the dirty laundry, and the poor project just needs some tender loving care to get back on track.  So how do you navigate these turbulent waters?

Simple.  Be specific when writing the reasons, impacts, and corrective actions.  Do not add unnecessary information or lay blame—just stick to the facts.  Identify the tasks behind schedule, provide a brief reason, and then focus on the impact to the project and the corrective actions being taken.  And the number one mistake I see in schedule variance analysis is confusing cost with schedule.  For example, if any part of the explanation for the variance has the word “cost” in it, you’re headed in the wrong direction.  An overrun in cost is not even part of the schedule variance equation!  You’re only looking at how much work was scheduled compared to what was done.  Also, remember that you cannot take earned value performance for work that is not in the baseline. 

Sometimes poor variance write-ups are due to missing the real cause of the variance.  Finding the correct cause can often lead to the appropriate impact and corrective action.  So, if you’re really struggling with identifying the cause of the variance, remember, possible schedule variance (SV) causes include:

• Poor baseline schedule (does it reflect reality?)
• Subcontractor/vendor cannot deliver when needed
• More/less effort than planned (be specific)
• Insufficient resources (staffing)
• Labor disputes/work stoppage,
• Resource availability (is it there when I need it?)
• Requirement changes 

All of the above examples are issues that can directly affect the earned value realized by the project.  If you have gone to the trouble to implement the elements of EVMS, at least be sure to maintain its integrity with strong schedule variance analysis.

This is part one of a two part series on variance analysis.   Part Two takes a look at some real world examples and examine why they are good or bad variance write-ups.   

Click here to view : Part Two of “How to avoid being sucked into the vortex of Schedule Variance Analysis”

Next week we’ll tackle the CPI report and look at how to write a solid cost variance analysis. 

- Melissa Duncan (About Melissa)

Read the previous posting Earned Value Exposed: What has Your Earned Value Software Done for You Lately?

Subscribe to Blog 

Earned Value legend as there may be a few of us that don’t yet have this memorized…

PM=Project Manager

PC=Project Controls

CAM=Control Account Manager
CPI=Cost Performance Index

EVM=Earned Value Management

EVMS=Earned Value Management System
WBS=Work Breakdown Structure

OBS=Organizational Breakdown Structure
SPI=Schedule Performance Index
IPMR=Integrated Program Management Report

How to avoid being sucked into the vortex of Schedule Variance Analysis: Part Two

The discussion in Part One reminded us to stay focused on the facts as an essential aspect of completing schedule variance analysis in support of earned value management, with a few tips on what not to do.  Let’s take a look at some examples and examine why they are good or bad variance write-ups.  These are real examples from real projects that were behind schedule from an earned value perspective (earned value was less than scheduled/planned value).

Good Schedule Variance Report – Example #1 
Reason: Progress slowed due to difficulty of removing control rod drive mechanisms.  Shipping of CRDM’s to laboratory was behind schedule due to laboratory documentation and procedural issues.  Non-Destructive Assay (NDA) is slow due to resources working on higher priority lab projects.
Impact: No long-term impact to the project with additional labor support for NDA coming in February.  With proper support the project will be back on schedule in March.
Corrective Action: Continue to use selective overtime and make adjustments in the sequencing of work.  Working with the laboratory to resolve documentation issues and NDA labor availability.
Comments:

• The reason statement is clear on what work has been delayed and why, yet it is very brief.
• The impact statement tells the possible results/severity of the delay and addresses when recovery will be expected (including the “with proper support” caveat condition).
• Action statements tell what is being done to recover the slip in schedule.

Good Schedule Variance Report – Example #2 
Reason: Behind schedule on removal of TR-N1 due to equipment breakdowns and poor-weather delays (excessive winds).
Impact: TR-N1 removal is three weeks behind schedule; which, if not recovered, will impact replacement schedule of new transformer and the testing/turnover milestone.
Corrective Action: A recovery schedule has been prepared and implemented that will recover the schedule in time to meet the testing/turnover milestone.  Equipment repairs have been made but schedule recovery is based on continuous equipment operation.  Routine maintenance frequency has been increased to minimize the extended downtime.  Additional spare parts have also been purchased.  With the seasonal change, there should not be weather delays during the recovery period.
Comments:

• The reason/issue is brief but it is clear what caused the delay and why there is a schedule variance.
• The impact addresses that a milestone will be affected if action isn’t taken.  This provides an early warning, documents the current situation, and will get management support for recovery actions.  The specific milestone impacted is also stated.
• The corrective action statements tell what is being done to recover the slip in schedule and the expected result (when).  Additional information is added to help clarify the expected results.

Good Schedule Variance Report – Example #3
Reason: Extended discussions by the utility administration and the board of directors led to delays in starting the following work: Task #1 (name the task)
-     Task #2 (name the task)
       -     Task #3 (name the task)
Impact: The delay will not allow completion of task #2 this fiscal year, as planned.  Tasks #1 and #3 can be completed this FY but with revised completion dates.  The milestone deliverable dates cannot be met.
Corrective Action: A revised schedule is being prepared to incorporate the later start dates.  Since the delay is beyond company control, a baseline change will be prepared to incorporate the budget and resource changes into the Baseline.  This change will not have a significant impact to overall plant operations.
Comments:

• The reason statement specifies the tasks that are impacted by the delay and why.
• The impact statement sates the expected impact to the tasks specified (unrecoverable) and that a key milestone is also impacted.
• Corrective actions relate to what will be done relative to the impact.

Poor Schedule Variance Report - Example #4
Reason: Work was behind schedule at the end of the month.
Impact: Negative schedule variance.
Corrective Action: Full recovery will be achieved.
Comments:

• Reason needs to tell the reader what caused the delay and why.
• The impact statement needs to assess the impact on other work, key milestones, and the activity completion date. 
• The corrective action needs to identify what the project is doing to correct the situation.

Poor Schedule Variance Report – Example #5 
Reason: Behind schedule due to lack of resource availability
Impact: Expect to recover schedule.
Corrective Action: Offers have been extended.
Comments:

• The reason needs to be more specific; it should address the work that is being delayed.  Why adequate resources were not provided should also be addressed. 
• The impact statement does identify the expected result but it also needs to state the impact on other work or key milestones.  It’s appropriate to state that the work is not on the critical path and there is no impact to milestones.
• The corrective action should be more specific in identifying what the project is doing to correct the situation; i.e., how many and what personnel are being added and when is recovery expected?  Usually, overtime is used rather than a new hire; if so, consider the cost variance impact as well.

Writing a proper schedule variance analysis does not have to be difficult, but you will need to spend enough time to think carefully and ensure you have addressed the specific issues adequately.  Do not try to rush through the write-ups or you risk undermining the benefits of earned value management.  Good luck!

Do you have any great examples of what to write or not write?  I know some of you must have some interesting stories, so consider sharing them with us.  Please share in the comments below. 

Next week we’ll tackle the CPI report and look at how to write a solid cost variance analysis.

- Melissa Duncan (About Melissa)

Read the previous posting Earned Value Exposed: What has Your Earned Value Software Done for You Lately?

Subscribe to Blog  

Earned Value legend as there may be a few of us that don’t yet have this memorized…

PM=Project Manager

PC=Project Controls

CAM=Control Account Manager
CPI=Cost Performance Index

EVM=Earned Value Management

EVMS=Earned Value Management System
WBS=Work Breakdown Structure

OBS=Organizational Breakdown Structure
SPI=Schedule Performance Index
IPMR=Integrated Program Management Report

What Has Your Earned Value Software Done For You Lately?

Have you ever cursed your Earned Value management software for not easily giving you the data you want? Perhaps it does not provide certain reports, thus necessitating the use of additional software just to get the data out in a palatable format.  Or maybe you are always pushing the data into another software just so you can analyze the problems with this month’s performance. As sad as it sounds, it is rather common to use one software to generate data, and yet another to report or analyze it.  For example, raise your hand if your Earned Value software can generate ready-to-publish CPRs or IPMRs without using another software program or add-on.  It’s unlikely you raised your hand.  Now I want you to ask your EVM software, “What have you done for me lately?”  Perhaps when Eddie Murphy and Janet Jackson made those words famous, they weren't referring to EVM software.  However, it doesn't mean you can’t apply the concept to it!  

Why are we doing this to ourselves? Well, in simple terms, it’s because we haven’t had much choice.  EVM software typically comes with a few canned reports, and most are clunky when it comes to drilling down into your 
data or slicing and dicing it.  Sometimes a user must look at more than one screen in the software to put all the pieces together to find a problem.  Let’s be honest; drilling down and finding exactly the activity/resource that caused the funky variance should not require three monitors so you can find the problem.  It’s time consuming and costly—especially if you’re forced to trace the issue through multiple software programs.  Don’t forget to add in the headaches associated with buying added software, creating the interfaces, “massaging” the data to load it into the next software, oh, and the added manual steps that are just begging for a mistake!  This isn't exactly the automation dream you were looking for, is it?  

EV software is supposed to make your life easier, by highlighting areas of concern early enough to make positive changes for the project.  The valuable EV data within the system absolutely must be easy to access, otherwise one of two things can occur: you risk having a “surface level” EV solution that won’t be used for real analysis and control of the project.  Or, you may be spending too much time getting the data out of your EVM software for reporting and analyzing—which, let’s face it, is wasteful.  Your time would be better spent actually trying to correct the problems on the project, not fight with your EVM software.

Setting up your structure or data inputs with the end goal in mind can help make the canned reports more meaningful and useful.  But if you’re already underway and changing the inputs would be too difficult, then what recourse do you have?  You’re forced to change the outputs.  That usually means added software, whether it be wInsight, Empower, MS Excel, MS Access, or sometimes a combination of several software programs.  And don’t forget what happens when a new manager comes in and they want to see the data sliced in a different way.  Might as well attempt to pull a rabbit out of your hat.  So do your research and choose an EVM software that really can save you time and money while giving you the data you need.  Software that integrates cost and schedule can be pricey, so of course, being prudent that you’re getting the biggest bang for your buck only makes sense. A good EVM application will have ready-to-publish reporting capabilities with solid drill-down functions as standard features.   Limitations in any software is a given, but if you do your research, you CAN find an EV software that will generate the data in a format you can actually use, without forcing you to buy additional software.  

To aid you in your quest for better earned value management software, below are some things a serious earned value management system ought to be able to easily perform or address:

• Generating pre-formatted and ready to publish CPR/IPMRs, SPA and CPI/SPI graphs, cost element pie charts, Control Account Plans, dashboards, etc., at various levels in the WBS and filter to exclude overheads, G&A, etc.  You shouldn’t need another software program to generate the necessary reports.
• Drill-down capabilities, such as clicking on a control account and automatically drill-down into the work packages, control account plans, variances, or cost elements.
• Definable variance thresholds at any level of the WBS and automatic flagging when a threshold is exceeded, along with a screen to enter the variance analysis and corrective actions to produce your variance reports (CPR/IPMR Format 5).
• Generating pivot tables within the software, instead of having to export it to MS Excel or MS Access.
• Speaking of exporting, your EV software should be able to export data, easily.  Only one or two clicks should be needed to get the data out of the software.
• If you do need to push the data into another program, your EV software should at least have open integration with other software.  It should be capable of utilizing simple interfaces.

Is there anything else you would add to this list?  How do you feel about your EVM software? What do you WISH it could do?  Please share your input in the comments.

Our next post looks at avoiding the political storm caused by weak schedule variance analysis. 

- Melissa Duncan (About Melissa)

Read the previous posting Earned Value Exposed: Obtaining EVMS Compliance When Management Can't Even Spell EVM.  

Subscribe to Blog