The Spider Team Reference Library

SUCCESS DRIVEN PROJECT MANAGEMENT Vladimir Liberzon, Spider Management Technologies, general director

1. Introduction In this paper we will discuss the proven methodology of project planning, performance analysis and project control that we call Success Driven Project Management (SDPM). This methodology integrates scope, time, cost, and risk management and may be of particular interest for the Critical Chain theory supporters because both approaches have some common features. We will also discuss some notions and methods of project planning and performance analysis that are common in Russia and necessary for SDPM understanding. We will illustrate the approaches described in this presentation using project management software package Spider Project that is most popular professional PM tool in Russia.

2. Resource Critical Path Definition We shall start with the definitions and the first of them is the definition of the Critical Path. We use the term Resource Critical Path (RCP) to specify our interpretation of the classical PMBOK Guide® definition. We believe that project Critical Path, Resource Critical Path and Critical Chain: a) imply the same set of activities and b) the traditional interpretation of the critical path is not correct. A Guide to the Project Management Body of Knowledge  defines the Critical Path as those activities with float less than or equal to a specified value, usually zero. Float is the amount of time that an activity may be delayed from its early start without delaying the project finish date. Early start is the earliest possible point in time at which the uncompleted portions of an activity (or the project) can start, based on the network logic and any schedule constraints. Project schedule constraints include resource constraints, finance and supply constraints, calendar constraints and imposed dates. The float should be calculated with all schedule constraints as well as the network logic taken into account. The total float determined by most PM packages shows the time reserve for the execution of activity, however the availability of resources is completely ignored. It is not the actual activity float as defined by A PMBOK Guide. True critical path should account for all schedule constraints including resource and financial limitations. We call it Resource Critical Path (RCP) to distinguish it from the traditional interpretation of the critical path definition. The calculation of RCP is similar to the calculation of the traditional critical path with the exception that both the early and the late dates are calculated during forward

http://creativecommons.org/licenses/by-sa/3.0/

(1)

The Spider Team Reference Library

and backward resource (and material, and cost) leveling. It appears that by adding financial and supply constraints to the Critical Chain definition as well as the way of the Critical Chain calculation, we will obtain something very similar to RCP. RCP can consist of activities that are not linked to each other. Traditional critical path approach assumes that this may be due to the different activity calendars and imposed dates. In case of RCP calculation, it can also be due to resource constraints and financial and supply limitations.

3. Activity Volume and Resource Productivity Projects are often planned (especially in the construction and manufacturing) basing on the federal, local, industrial or corporate norms and standards. These norms usually refer to resource productivity on the certain activity types, costs and materials per unit of activity volume (volume of work to be done on activity). Usage of these norms affects the planning of project activities. Activity volume can be measured in meters, tons, etc., planned work hours, percents or any other units. It is often used as initial activity information instead of duration. If assigned resource productivity is defined in volume units per hour then activity duration may be calculated during project scheduling. Activity volume does not depend on assigned resources. Calculation of activity duration basing on assigned resource productivity has many advantages. We have already mentioned the possibility of applying corporate norms. Changing the expected resource productivity we change the planned duration of all activities of certain types. It is especially useful for the forecasting of project duration and estimating uncertainties.

4. Resource Analysis Resource analysis is an essential part of project execution analysis. It is vital for the project time analysis to be able to forecast resource productivity. Monitoring of the actual resource performance allows to determine trends and to make necessary adjustments of resource productivity databases. The usage of project resources varies at the different project phases. The forecasting that accounts for these differences is considerably more accurate than the methods of Earned Value Analysis. Methods of risk analysis and simulation should include estimations and simulations of activity volumes, resource productivity, resource availability, etc. The main problem with the traditional methods of risk simulation is their initial assumption that cost and duration deviations of different activities are independent of each other. Activities performed by the same set of resources will have correlated duration. Not taking this into consideration leads to producing wrong risk simulation results.

http://creativecommons.org/licenses/by-sa/3.0/

(2)

The Spider Team Reference Library

5. Sample Project Let’s illustrate the basic concepts described earlier using sample project consisting of only three independent activities and two resources. Project data are shown in the Table 1. Activity Name

Activity Volume

Assigned Resource

120 CE

ACTIVITY 1

Resource Productivity

Resource Hour Cost

0.5

R ESOUR 1

50

ACTIVITY 2

60

R ESOURCE 2

0.3

30

ACTIVITY 3

60

R ESOURCE 2

0.3

30

Table 1. Sample Project Data Project schedule before resource leveling is shown at Figure 1. Activity 1 is critical. Other activities have 5 day float.

Figure 1. Sample project schedule before resource leveling After resource leveling activities 2 and 3 became resource critical while activity 1 has 20 day float (resource float). So RCP consists of activities 2 and 3 (Figure 2).

Figure 2. Sample project schedule after resource leveling

http://creativecommons.org/licenses/by-sa/3.0/

(3)

The Spider Team Reference Library

Activity resource floats have one large advantage over the total floats calculated by most PM software. This advantage is feasibility. Traditional total float shows the period for which activity execution may be postponed if project resources are unlimited. Activity resource float shows the period for which activity execution may be postponed within the current schedule with the set of resources available in this project.

6. SDPM Project Planning Technology Our experience of project planning shows that the probability of successful implementation of deterministic project schedules and budgets is very low. Therefore project planning technology should always include risk simulation to produce reliable results. We will describe the approach to project planning and risk simulation that is supported by Spider Project. The project planner obtains three estimates (optimistic, most probable and pessimistic) for all initial project data. These data are used to calculate optimistic, most probable and pessimistic project schedules and budgets. The most probable and pessimistic project versions will usually contain additional activities and costs and employ other resources and different calendars than the optimistic schedule. The planner should define desirable probabilities of meeting target dates, costs, and material consumption rates at main elements of project WBS. Basing on these probabilities, the package calculates corresponding desired project target dates, costs, and material requirements. These desired data form the basis for contract negotiations and decision making. Negotiations may result in establishing new target data. Project risk simulation results help to negotiate by answering the questions on probability to meet any restrictions on time and on budget. Probability of meeting target data (cost, time, quantity) is called Success Probability. Success Probability is the BEST indicator of the current project status. In addition, the package calculates the target schedule. Target Schedule is the backward project resource constrained schedule with the most probable activity duration, material requirements and costs and target dates of the project phase completion. Let’s apply the described approach to our sample project.

7. Sample Project Planning We will make very simple assumption: • •

Optimistic productivity of assigned resources is 20% higher and pessimistic is 20% lower than the planned ones. Let’s assume that we have no other risks :)).

http://creativecommons.org/licenses/by-sa/3.0/

(4)

The Spider Team Reference Library

Estimations of resource data are shown in the Table 2: Resource Name

Most Probable Resource Productivity

Optimistic Resource Productivity

Pessimistic Resource Productivity

Resource 1

0.5

0.6

0.4

Resource 2

0.3

0.36

0.24

Table 2. Estimations of Project Sample resource productivities We want to be on time with 70% probability and under budget with 75% probability. The package will identify the desired finish date and the required project budget: Parameter Name Finish Date

Most Probable Value 10.05.2002 16:00

Desired Probability 70

Project Budget

24000.00

75

Desired Value 15.05.2002 14:08 25612.90

Table 3. Project Sample desired data Next Figure shows Target schedule for the project Sample:

Figure 3. Project Sample Target schedule Let’s assume that after negotiating the contract we established project target finish date and target budget. The package will calculate the probability of meeting target parameters and we will see that we were lucky with time and should be very cautious with the project budget: Parameter Name Finish Date Project Budget

Target Value 15.05.2002 16:00 25000.00

Success Probability 72.10 63.71

Table 4. Initial Success probabilities for the project Sample

http://creativecommons.org/licenses/by-sa/3.0/

(5)

The Spider Team Reference Library

Let’s assume that the most probable version of our project was defined as the project baseline. The initial probabilities to meet most probable project parameters are shown in the Table 5: Parameter Name Finish Date Project Budget

Target Value 10.05.2002 16:00 24000.00

Success Probability 40.93 40.93

Table 5. Initial probabilities of meeting most probable data for the project Sample

8. Sample Project Execution Now let’s simulate Sample project execution. Let’s assume that the actual resource 1 productivity was 10% higher than expected (0.55), while the actual resource 2 productivity was 10% lower (0.27) and the productivity of these resources did not change during project execution. Our task - to analyze project performance, to forecast future project results, and to decide if corrective action is necessary. We will use Resource analysis, Success Probability analysis, and Earned Value analysis. We assume that estimates were done each week and will analyze trends. Initial baseline data (estimate at completion):  

Project Finish Project Cost

- 10.05.2002 16:00 - 24000.00

9. Resource Analysis 1st week performance analysis and adjustments: Providing Resource Analysis we shall decide if the actual deviations in resource productivity are accidental or the planned productivity should be adjusted. Let’s adjust them raising productivity of resource 1 by 5% (half of the difference between planned and actual productivity - to 0.525) and lowering the productivity of resource 2 by the same 5% (to 0.285). Our new forecast (estimate at completion):  

Project Finish Project Cost

- 15.05.2002 15:09 - 24028.09

http://creativecommons.org/licenses/by-sa/3.0/

(6)

The Spider Team Reference Library

Figure 4. 1st week forecast by resource analysis Advice: your project may be late and over budget 2nd week performance analysis and adjustments: New performance data show that the planned resource productivity should be adjusted again. Let’s adjust them raising productivity of resource 1 to 0.5375 and lowering the productivity of resource 2 to 0.2775). Our 2nd forecast (estimate at completion):  

Project Finish Project Cost

- 17.05.2002 10:35 - 24107.63

3rd week performance analysis and adjustments: New performance data show that the planned resource productivity should be made equal to the actual because they did not change from the project start. Our 3rd forecast (estimate at completion):  

Project Finish - 20.05.2002 12:26 Project Cost - 24242.44

Figure 5. 3rd week forecast by resource analysis

http://creativecommons.org/licenses/by-sa/3.0/

(7)

The Spider Team Reference Library

These forecasts will not change later. The forecasts based on resource analysis might be very accurate but are not easy to make and require detailed performance reports. Besides they do not consider project risks that depend on factors other than resource performance.

10. Success Probability Analysis Success Probability shows current project status, Success Probability trends show project manager if corrective action is needed. Success Probability trends for our Sample project is shown in the Figure 8.

Figure 6. Success probability trends for the project Sample We estimate probability of meeting target project finish date and target and baseline cost. Success probability trends show us that the project will be:   

certainly late and probably over budget of $24000 though certainly under budget of $25000.

The problems with the project timely performance were shown from the very beginning though problems with meeting baseline cost became obvious only after the actual finish of activity 1 execution. Success probability may change not only due to performance deviations but also because risk estimates were changed and it caused changes in pessimistic and most probable project versions.

http://creativecommons.org/licenses/by-sa/3.0/

(8)

The Spider Team Reference Library

11. Earned Value Analysis Earned Value Analysis applied to the project Sample showed the following trends: 0 1 1 2 0 0 1 2 2 Parameter 4.03.2002 1.03.2002 8.03.2002 5.03.2002 1.04.2002 8.04.2002 5.04.2002 2.04.2002 9.04.2002

ACWP

3200

6400

600

9 2800

1 6000

1

1 1 2 2 8109.1 9309.1 0509.1 1709.1

BCWP

3280

6560

9840

13120

16400

18480

19560

20640

21720

BCWS

3200

6400

9600

12800

16000

19200

20400

21600

22800

CV

80

160

240

320

400

370.89

250.89

130.89

10.89

CV%

2.5

2.5

2.5

2.5

2.5

2.05

1.3

0.64

0.05

SV

80

160

240

320

400

-720

-840

-960

-1080

SV%

2.5

2.5

2.5

2.5

2.5

-3.75

-4.12

-4.44

-4.74

CPI

1.02

1.02

1.02

1.02

1.02

1.02

1.01

1.01

1

CPI%

102.5

102.5

102.5

102.5

102.5

102.05

101.3

100.64

100.05

SPI

1.02

1.02

1.02

1.02

1.02

0.96

0.96

0.96

0.95

SPI%

102.5

102.5

102.5

102.5

102.5

96.25

95.88

95.56

95.26

Table 6. Earned Value Management data trends for the project Sample You may notice that the problems with the project Sample schedule and budget are hidden till the finish of activity 1 and even later. You should analyze trends to discover that the problems exist. It can be explained by the fact that EVA does not consider project logic dependencies and thus the successful execution of activities that don’t belong to the critical path may lead to the good EV parameters despite the poor forecast of project completion date and budget. We recommend to apply EVA to the critical path activities of the project life cycle phases to forecast the phase duration and budget. You also cannot use EV data from one phase to forecast the execution of another if the resources used on these phases are different.

12. SDPM Project Management Technology We recommend using the optimistic project version for setting tasks for project implementers while the calculated contingency reserves should be used by the PM team for the management purposes.

http://creativecommons.org/licenses/by-sa/3.0/

(9)

The Spider Team Reference Library

Start (finish) contingency reserves (buffers) are calculated as the difference between activity start (finish) time in the optimistic and target schedules. Contingency reserves are also calculated for the activity cost and material requirements. But the most valuable indicators of project performance are Success Probabilities - probabilities of meeting target project parameters. Trends of Success Probabilities show project manager if the corrective action is needed. The value of Success Probability shows current project status better than any other project parameter. Project manager obtains the following estimates necessary for effective project control:  

Probabilities of meeting target project parameters (success probabilities), Target activity start and finish dates, resource and material requirements and cost, Planned activity start and finish dates, resource and material requirements and cost in the current schedule, Activity resource floats that show the time for which activity execution may be postponed without delaying project finish date in the current schedule, Activity contingency reserves (buffers) for time, cost and materials calculated as the difference between the corresponding optimistic and target parameters.

  

Figure 7 shows the optimistic and target data, and other scheduling and risk analysis information for our Sample project.

Figure 7. Initial SDPM data for the project Sample SDPM tips for the project control: • • • •

Plan day-to-day activities using the optimistic estimates but pay special attention to resource floats and to contingency reserves. Include the causes of delays in activity completion and cost overruns in performance reports. Regularly update the estimates in the optimistic, most probable and pessimistic project schedules. Regularly recalculate the success probabilities and analyse trends.

13. SDPM and Critical Chain You may notice that SDPM and Critical Chain approaches have a lot in common:

http://creativecommons.org/licenses/by-sa/3.0/

( 10 )

The Spider Team Reference Library

• • • •

Resource Critical Path is the same as Critical Chain. Therefore Critical Chain project buffer may be regarded as an analogue of SDPM contingency time reserve, feeding buffers are similar to resource floats, Both the SDPM and Critical Chain approaches recommend to use the optimistic estimates for setting the tasks for project implementers.

Success probability trends show if project (phase) time and cost reserves were spent faster or slower than expected. It is more effective than not obvious estimates if Critical Chain project buffers were properly utilized. But there are differences too. We cannot agree with the Critical Chain theory’s assumption that one should always avoid multitasking. Usually there are many subcritical activities belonging to the different network paths and even the minor delays in the execution of subcritical activities can lead to the changes in the RCP. This comes into conflict with the Critical Chain theory’s assumption that the Critical Chain never changes during the project execution. The assumption that only one project drum (in our terminology - critical) resource exists is also dubious. Our experience shows that critical resources are different at the different phases of project lifecycle.

14. SDPM technology steps 1. Assess risks and create optimistic, pessimistic and expected estimations of activity duration and volumes, resource productivity and quantity, activity and resource cost, calendars, etc. Estimates are usually based not only on expert judgments but also on regulatory or corporate norms. 2. Create and calculate optimistic, most probable, and pessimistic resource and cost constrained project schedules. 3. Calculate project and each phase desired finish dates and costs (with the desired probability of successful execution). 4. Negotiate and define target finish dates and costs. 5. Calculate target schedule with the most probable activity duration but target finish dates (backward resource constrained schedule). 6. Determine time and cost buffers, and probabilities of meeting target parameters. 7. Use the optimistic schedule as a plan for the employees. Using optimistic estimates, we increase our chances to obtain the reports on any deviations from the proper activity execution. This information is necessary for quality and risk analysis. 8. Control risks and regularly recalculate project schedules and necessary time buffers and success probabilities. Success probability trends show if corrective actions are necessary.

http://creativecommons.org/licenses/by-sa/3.0/

( 11 )