TCPI

Abstract

This paper examines the To-Complete Performance Index (TCPI) as one of the forecasting tools of Earned Value Management (EVM). We explore why project personnel should care about Earned Value Management's To-Complete Performance Index (TCPI), what TCPI represents, how to calculate TCPI, and what the calculated TCPI results mean. A simple analogy for TCPI is described to set a conceptual understanding. A brief history of Earned Value Management is discussed. TCPI is traced to when it was introduced into project management as a cost control tool. TCPI is defined in words, formulas, examples, and what the results mean. One should walk away from this presentation with a comprehensive understanding of the power of TCPI and what it can do for you as a project stakeholder.

How many of you know what a “To-Complete Performance Index” (TCPI) is? Does anyone care? Who really wants to know? Why should anyone be interested in TCPI? What will TCPI tell us that we may want to know? How will TCPI help us control projects? All good questions! All of which we shall answer here.

All project managers, project team members, project sponsors, functional managers, senior managers, and other stakeholders should be interested in TCPI, especially for those projects in distress (cost) and other projects that might be headed for distress.

In a simple statement, TCPI answers the question of “how efficiently must we use our remaining resources (financial)?”

An analogy for TCPI:

You are driving your car to meet a friend at a movie theater. You told your friend you would meet him there by 7:00 pm. You live 30 miles away and left home at 6:00 pm. It is now 6:30 pm and you have only driven 10 miles. You have 20 miles to go (you have averaged 20 mph for the first 30 minutes due to traffic). What speed must you average for the last 20 miles in order to arrive on time? Using the TCPI method/formula, it would calculate out to 40 mph. This result describes the performance required in the last 20 miles of distance to meet the original preset target of 7:00 pm. The TCPI answer of 40 mph should cause you to think about how you are going to improve your average speed from 20 mph to 40 mph. Take a different route? Maybe! Call your friend and tell him you are going to be late? Again, maybe!

TCPI is a forecasting feature of earned value management. EVM has been around since the turn of the 20^th, century when the concept of “earned time” in industrial manufacturing was first introduced. EVM was further developed in the 1960s by the U.S. Department of Defense as part of their PERT (Program Evaluation and Review Technique)/COST Method.

Project Management Institute (PMI) first picked up on EVM as a project management cost control tool in the late 1980s in an early version of A Guide to the Project Management Body of Knowledge (PMBOK^® Guide). PMI has been testing Project Management Professional (PMP)^® candidates on EVM, but not the TCPI formula since the earliest days. Then TCPI and its explanation appeared in PMI's PMBOK^® Guide—Third Edition, but PMI still did not test on it. When the PMBOK^® Guide— Fourth Edition was released, PMI began testing on the TCPI formula and the results it predicts.

TCPI is a calculated projection of cost performance that a project must achieve on the value of the remainder of the project work to achieve a specified end result. The project end result could be the current budget at completion (BAC), estimate at completion (EAC) (current or new), or even a management specified end result goal. The goal could be higher, the same as, or lower than the current BAC or EAC.

TCPI for BAC is conceptually shown in Exhibit 1 below:

Exhibit 1 – TCPI for BAC.

TCPI for EAC_new (new ETC_new) is conceptually shown in Exhibit 2 below:

Exhibit 2 – TCPI for EAC.

Let's examine a project that is performing poorly, related to the budget. The current cost performance index (CPI_c) is probably 0.85 or less. You want to know what cost performance index (CPI) you must achieve on the value of the remaining work on the project to bring the project in on BAC.

The TCPI to bring the project in on the BAC is the ratio of the value of the remaining project work, per PMI's definition [BAC minus earned value (EV)], all divided by the amount of the remaining funds [BAC minus actual cost (AC)].

This formula works out to be:

TCPI = (BAC – EV) ÷ (BAC – AC).

(BAC – EV) is the value of the remaining work, per PMI's definition.

(BAC – AC) is the amount of funds remaining.

The lower the initial CPIc, the higher the TCPI will calculate. The later in the project this evaluation is performed, the higher the TCPI will calculate.

For all CPI_c less than 1.0, the TCPI will be greater than 1.0. When TCPI turns out to be greater than one (> 1.0), a more normal case for BAC calculations, the value of the remaining project work must be executed at a better cost performance level than the project's completed work.

In another situation, if TCPI turns out to be one (1.0), a somewhat unusual case (project on budget), the value of the remaining project work must be executed at the same cost performance level as the previously completed project work. This may or may not be difficult, depending on the nature of the remaining project work.

On a given project, once it becomes clear that the current BAC is not obtainable, the project manager/team, with sponsor approval, should develop a new forecast EAC by completely re-estimating the cost of the remaining work (new bottoms up ETC). The new EAC becomes AC + ETC_new. Once the new EAC is approved by the project sponsor, the new EAC replaces BAC in the TCPI formula denominator and the TCPI formula becomes:

TCPI = (BAC – EV) ÷ (EAC – AC).

Analysis of the two TCPI formulas above, shows when using BAC, it is effectively the budget for the remaining work, as defined by PMI (BAC – EV), all divided by the remaining funds (BAC – AC). TCPI, when using EAC, is still the budget for the remaining work (BAC – EV), all divided by the remaining newly approved project funds (EAC – AC).

The next issue to face the project manager and/or project sponsor with TCPI results is when TCPI does in fact calculate to be greater than CPI_c. The question that must be answered is: “Is this new level of cost performance actually realistically obtainable?” The real answer to this question probably depends on a number of factors, such as:

1. What are the remaining project risks?
2. How tight is the remaining schedule?
3. What are the remaining resources?
4. What is the technical nature of the remaining work?
5. Are there any quality issues?
6. Any number of other factors, depending on the type and nature of the project.

Even when one knows the answers to the above questions, it really is probably still a judgment call. Surely we all remember that cost overruns can be improved, but usually are not fully recoverable.

TCPI's usefulness is best illustrated by a scenario. There are four possible scenarios:

Scenario number one: Current CPI_c is greater than one. TCPI will calculate out to be less than one. Not a lot of interest here.

Scenario number two: Current CPI_c exactly one. TCPI will calculate out to be exactly one. Not a lot of interest here.

Scenario number three: Current CPI_c is less than one, but in the 0.9 range. We shall examine TCPI for both the BAC and EACnew cases (3A and 3B).

Scenario number four: Current CPI_c is much less than one, maybe in the 0.7 range. There is no need to examine TCPI for the BAC case. It is just not doable, becauseTCPI will be significantly greater than 1.0, and probably in the 1.2 to 1.4 range. The actual number depends on how complete the project is when the calculation is performed. TCPI for the EAC_new case is the same approach as outline in scenario 3B, above. No need to repeat here.

So, let's start with scenario 3A:

Scenario number 3A: Let's consider a simple example to see how the numbers work.

Project data:

BAC = US$575,000

EV = US$230,000

AC = US$252,000

The project is approximately 40% complete (EV/BAC).

THE QUESTION: What TCPI is required to bring this project in on the current budget (BAC)?

First, let's calculate the current CPI_c.

CPI_c = EV/AC = US$230,000/US$252,000 = 0.913

Now, let's calculate TCPI.

This means the value of the remaining work on the project must be executed with a 1.068 cost performance level to bring the project in at the current BAC. The question that remains is “considering the past cost performance index has been 0.913, how likely will it be that it can be raised to 1.068 (a 17% improvement)?” In this case, somewhat unlikely, I would think!

Scenario number 3B: Let's consider another example, but just a bit more complex.

Project data, the same as above, except: a new bottoms up ETC_new of US$350,000 has been approved by the project sponsor [EAC is now US$602,000 (AC + ETC) (US$252,000 + US$350,000)].

First, let's calculate the current CPI_c.

CPI_c = EV/AC = US$230,000/US$252,000 = 0.913, same as before.

No surprise here I hope.

Now, let's calculate TCPI using the newly approved EAC.

Was the 0.986 answer surprising? Did you expect it to be something else? Let's see why TCPI is not some other number. Per PMI's definition, the value of the remaining work remained at US$345,000. The remaining funds increased from US$323,000 to US$350,000, an increase of US$27,000. TCPI came down, because the value of the remaining work remained constant (numerator) while the remaining funds (denominator) increased. Is it likely that the project team can raise cost performance on the remaining work from 0.913 to 0.986 (an 8.0% improvement)? Let's see the project manager's action plan.

The project manager's proposed action plan is:

1. Replace two inexperienced software coders with more experienced ones, saving US$5000.
2. New hardware has come available that will reduce cost by US$10,000.
3. The project manager is going to take on the additional duty of updating the schedule instead of using the PMO scheduler, which is a charged cost to the project, saving US$5000.
4. The project manager is going to cut the project team lunch hour from 60 minutes to 30 minutes, but provide catered lunches, saving US$7000.

Surely a discussion will ensue during which the action plan, its possibilities, and likelihood that the action plan will work will be discussed.

At the end of the day, it is probably still a judgment call. But, it will be much more likely than a TCPI of 1.068 or greater.

Three potential deficiencies of TCPI and its results are:

1. On a constrained end date project, TCPI fails to take into account the effect the schedule slippage has on the current and future CPIc and TCPI calculations.
2. Another point of concern is the term”the value of the remaining work” for the “EAC_new version of TCPI.” Is the value of the remaining work still the budget less earned (BAC – EV), as PMI suggests, or is it the new, approved ETC_new? One might argue that the “value” of the remaining work in the new approved ETC/EAC model has changed.
3. EVM and TCPI tell us about scope delivery, schedule, and budget. But neither have provisions and measurements for quality, risk, or resources. Thus, it is possible for EVM and TCPI to tell us one thing and entirely the opposite is true.

It is just food for thought!

TCPI is a great new tool for project people (sponsor, project management office, project manager, team members, and other stakeholders). TCPI allows you to calculate the amount of cost performance improvement that must be made on the value of the remaining work to reach a set goal (BAC, EAC_new, or a set management goal).

Glossary of Terms

1. AC – Actual cost (of work performed)
2. BAC – Budget at Completion (the sum of all budgets allocated to the project plus all approved change requests)
3. CPI – Cost Performance Index (the cost efficiency factor for a period of measurement)
4. CPI_c – Cost Performance Index Cumulative (the cost efficiency factor since the project began to the measurement period)
5. EAC – Estimate at Completion (a forecasted estimate of project cost when the project is complete)
6. ETC – Estimate to Complete (the estimated cost to perform the remaining work on a project)
7. EV – Earned Value (the value of the work completed)
8. EVM – Earned Value Management (a methodology of using PV, EV, and AC to determine project status and to forecast the end result)
9. PERT – Program Evaluation and Review Technique (a method to analyze the time needed to complete project tasks and identify the minimum time needed to finish a given project)
10. PMBOK^® Guide – PMI's A Guide to the Project Management Body of Knowledge
11. PMO – Project Management Office