The Handbook of Project Management: A Practical Guide to Effective Policies and Procedures, 2nd Revised Edition_7 pptx

Accuracy: ± 35–50% Project selection – top–down estimate This estimate is used to prepare a proposal and business case for aproposed project to seek PST approval and pass through Phase G

When third parties are required to carry out part of the work of a project

it is essential to come to an understanding with them about the basis oftheir estimates Remember that their estimate is probably a bid andincludes contingencies and profit

Contingencies are discussed later, but always avoid including gencies in your estimates Clearly identify extra time that is added for holi-days, etc However, if you are not careful people will add extra time andcost as a form of protection and argue that this can help bring in theproject early and under budget Ruthlessly strip such contingencies out ofestimates Padding estimates in this way can make an apparently goodinvestment look like a bad investment or, if the work is approved, with-hold funding and approval from other potentially valuable projects.These problems should never occur if you have an active PST and adoptthe principles and processes we have discussed in earlier chapters

contin-THE GOLDEN RULES

To minimize the probability of falling into the pitfalls pay particular tion to the golden rules of estimating:

atten-• The individual doing the estimating must have a good understanding

of the work he or she is estimating

• Anyone expected to be involved in actually doing the work must beinvolved in estimating the task durations These are the only peoplewho know their limitations (although they may not always admit tothem) and they will clearly identify when there is a steep learningcurve before output is matched to the schedule requirements

• Everyone involved in estimating must clearly understand the purposeand objectives of what they are doing and how to estimate effort andthe size of the tasks Only then will they produce estimates that willultimately come true and avoid the mistake of predicting their bestpossible performance

EFFORT AND DURATION

Fundamentally, an estimate is a decision about how much time andresource is required to carry out a piece of work to acceptable standards ofperformance This requires you to determine:

• The ‘size’ of the task or group of tasks, as determined from ments if possible

measure-• The amount of ‘effort’ required to complete the work How can thework be broken down? Can it be divided between two or more people?Effort is measured in project time units – hours, days or weeks.

• The level of risk involved

Once the effort is known it is possible to optimize the resource needs,taking individual capacities or available time into account to determine thelevels of effort required from each

Effort is a direct measure of a person’s time to do a piece of work innormal workdays Unfortunately that person will often have other non-project activities to complete, which reduces his or her capacity to do theproject work At a capacity of 50 per cent the work will take at least doublethe estimated number of workdays In practice it takes up to 20 per centlonger than this because of the ‘back-track’ effect due to the breaks in thecontinuity of the work Effort is measurable as continuous work with nointerruptions

Duration is a conversion of effort taking into account the number ofpeople involved, their capacities and an allowance for non-productivetime It is often more convenient to use full-time equivalents (FTEs) inderiving how many people are needed for any part of a project

Since duration is measured in real working days this is never the same

as the schedule, which has to take into account:

• non-available days for project work;

• non-working days – weekends;

• public and organization holidays;

• staff holidays

The first step for you is to derive some realistic durations and then applythese to a calendar to derive a schedule

Figure 7.5 The relationship between effort, duration and schedule

ESTIMATING THE DURATIONS

Of course everyone wants accurate estimates as the consequences of poorestimating are serious and expensive However, it is not practical in mostbusiness environments to spend a lot of time in detailed planning and esti-mating to derive the initial business case we discussed in Chapter 5 But

we do need the detail for the full plan development after PST approval So

we can really consider three levels of estimating, as described below

Idea evaluation – outline or order of magnitude estimate

This is the estimating process to avoid if you can Unfortunately there areusually more ideas seeking funding than funds available The outline esti-mate is an expert appraisal of predicted cost of the whole project forpresentation to the potential sponsor and seeks to pass through the initialscreening process discussed in Chapter 3 To save time it is frequentlybased on past projects using a scale factor The purpose of this estimate isonly to allow an evaluation and a decision to be taken about progressing to

a more detailed investigation of the potential project for submission to thePST This allows the idea to pass through Phase Gate Zero

Accuracy: ± 35–50%

Project selection – top–down estimate

This estimate is used to prepare a proposal and business case for aproposed project to seek PST approval and pass through Phase Gate One.This type of estimate uses past experience of similar projects where avail-able and has the goal of estimating the time and cost of completing eachkey stage identified in the work breakdown structure for the proposedproject Here it is essential for the golden rules of estimating to be appliedbecause you do not have the luxury of time to engage in detailed analysis

of each key stage The accuracy must be good enough to create confidence

in preparing a full business case and high level schedule for presentation

to the PST for approval It is common for feasibility studies to include thislevel of estimating

Accuracy: ± 15–25%

Project approved – detailed or bottom-up estimate

Once a project has been approved and taken through the definition phasethe PST makes a decision to open Phase Gate Two into the planning phase.Now the duration and effort for all the tasks in every key stage are anal-ysed in detail to derive accurate estimates for the schedule preparation

This estimate is based on all available data and must use information onresource capacity and availability from organization resource mapping.Bottom-up estimating is the most time-consuming but also the mostaccurate form of estimating The data for all the tasks can be rolled up foreach key stage using the work breakdown structure, allowing the prepara-tion of the key stage cost data and the project budget.

Accuracy: ± 5%

Sources of data

As the duration of each key stage is the real time it will take to complete the

work this is usually the most difficult part of the planning process.Unfortunately there is an abundance of ‘good advice’ in most organiza-tions about how long a piece of work will or should take The processappears to be part art and part science, which is hardly surprising sinceyou are really trying to predict the future! So far no one has produced areliable crystal ball!

The sources for accurate estimates are limited They are:

• experience of others;

• the expert view;

• historical data from other projects

There is no substitute for experience If similar work has been done before,you can ask others for their own experience and adjust the data for yourproject It is a reasonable way to start but always take a cautious approach.Elegant technical tools such as Monte Carlo exist for estimating but theydepend on a large database of information and are difficult to justify inmany situations

The data you collect in this way will often hide important relevant mation No one will easily admit taking longer than the plan predicted for

infor-a piece of work – pinfor-articulinfor-arly if slippinfor-ages cinfor-aused problems Also, people’smemories have a habit of only remembering the good news with thepassage of time If good plans and records exist review these to determinewhat actually happened compared to what was planned to happen Takemore than one opinion if you can and remember no two people ever dothe same piece of work at the same pace The equation relating effort andperformance is different for us all

Who are the experts? There may be a few – or so they believe! Alwaysask questions about how reality compared with original estimates forsome work Check that the nature or content of the work did not change.You soon discover who is above average at estimating accurately – theseindividuals are the experts you desperately seek Since it is relatively rare

for work to be identical between different projects, apply your adjusting

factor to arrive at a realistic duration for each of the activities Keep a record

of how you derived the estimates in case you are wrong, then you canimprove your estimating skills

People problems

Ask anyone how long a piece of work will take and you are likely to begiven a shrug and a smile and a wildly inaccurate answer This is becausepeople do not ask themselves some simple questions:

• Do I really understand what is involved?

• Do I have all the necessary skills and tools for the work?

• What else must I do at the same time?

• What is the priority of the project work compared with other work?

• When is the project work really needed by?

• Can I break the job down into chunks to do at different times?

• Can I predict what I will be doing when this project work needs to bedone?

• Will I be taking any holiday during the time concerned?

• Do I have any other obligatory commitments during the timeconcerned?

• What does my manager know about my future commitments that I donot know yet?

The reality is that the majority of people are not productive 100 per cent ofthe time! As much as 20 per cent of people’s working week is taken up by:

• meetings – particularly ones they need not attend;

• general interruptions:

– visits to desk and others – wanderlust;

– equipment failure;

– reading journals and e-mail;

– searching for information;

– giving support and advice to others;

• commitments to routine functional work and other projects;

Consider also:

• project complexity:

– specifications: adequacy, unfamiliarity;

– new quality standards;

– unclear understanding of the technology;

– new technology, which always has a learning curve for dence;

confi-• team size and location of the team members;

• anything else you can think of

The answers to the questions above and other similar questions are oftenignored in deriving an initial estimate, leading to considerable problemslater Rash promises are assumed as realistic and inserted into the plan

objec-then the contingencies are rolled forward and adjusted Ask:

• What factor can you use for adjusting people’s estimates?

• Is that factor global for all estimates or different for different types ofwork and for different people?

• Should you expose your adjustment factors?

• What limits must you use in applying contingency?

• Should you multiply some estimates by an additional weighting for:– team size;

– team experience (in individuals);

– team-working history of this team;

– project complexity;

– project use of new techniques or technology?

• If you think you should apply an additional weighting, what should itbe?

Use the answers to these questions to revise the base estimates A scope

contingency may be required if there is uncertainty about the detail of some

parts of the plan The scope contingency is only used when you as the

project manager specifically authorize its use through the change

manage-ment process Remember that you must have clearly delegated authority to

use the contingency

You take the final decision about the durations entered into the plan tocalculate the total project time with a projected completion date.Obviously there is a balance between the desired project completion dateand the projected or forecast completion date based only on estimates Theformer may well appear to be almost impossible and quite unrealistic, thelatter insupportable and good justification for cancelling the project!Somewhere in the middle there is an acceptable solution Always recordany assumptions you have made in deriving the durations.

‘phase’ or set of sub-projects that are then subject to detailed planning.This can lead to a stop–go situation with considerable uncertainty aboutresource availability and capacity and an unknown project completiondate – potentially the endless project!

The phase gate approach recommended in this book allows you to mate in detail at the key stage level and confidently derive a detailedschedule for total project cost and time computation The detail that needs

esti-to be exposed within each key stage is derived and validated as you movethrough the project work As issues arise and are resolved the impact onthe schedule can be readily evaluated with appropriate corrective actionagreed and undertaken The whole process has visibility to the stakehold-ers and significant changes can be proposed and agreed in the full under-standing of the impact on completion date and objectives

TIME-LIMITED SCHEDULNG AND ESTIMATES

There is always a conflict when a completion date is imposed on a projectbefore any work on estimates is carried out This imposed date iscompletely outside your control so you attempt to compress estimates tofit the plan To a limited degree this process is acceptable, in that it gives atarget, but too often it moves you into a totally unreal situation where youare faced with ‘mission impossible’ Apart from the complexity of the work

impacting the time element you have to make huge assumptions aboutresource availability You must still prepare realistic estimates to derive aclear case and state:

• what you can deliver in the time;

• what you cannot deliver in the time;

• why you can only meet part of the objectives of the project

You can then use your skill as a negotiator to arrive at an agreed solution!

Practical estimating

Since the major portion of all project costs is frequently the time expended,the accuracy of estimates is a key factor in achieving project success.Historical data, even from parts of previous projects, can be valuable as aninitial guide Analysis of the actual proposed work is essential if accurateestimates are to be derived Even then people seem to habitually underes-timate time for execution of work There are some people who have anintuitive ability to visualize the work involved and give accurate estimates– identify these people and make good use of them!

Some practical guidelines that can be used are as follows:

• Schedule full-time team members at 3.5–4.0 working (productive)days per week (to allow for holidays, absences, training courses, etc)

• Include management time where appropriate as an additional 10 percent

• In planning, avoid splitting tasks between individuals

• When tasks are split between two individuals do not reduce time by 50per cent – allow time for communication and co-ordination

• Take individual experience and ability into account

• Allow time for cross-functional data transfer and responses

• Build in time for unscheduled urgent tasks arising on other project activities

non-• Build in spare time for problem solving and project meetings

• Include appropriate contingencies at all levels of planning.

Any estimate is only as good as the data upon which it is based so, likeproject risks, accept that the estimate may change with time as more databecomes available to you As the project continues always review and vali-date the durations you have used For each key stage keep a record of:

• the estimates you have finally decided;

• any assumptions made during estimating;

• where contingencies have been added;

• how much contingency has been added

Now, with durations agreed, you can analyse the logic diagram for itscritical path.

IDENTIFYING THE CRITICAL PATH OF YOUR PROJECT

Critical path techniques have been in use on projects now for some 30years, having proved their value as a tool for project scheduling andcontrol The fundamental purpose is to enable you to find the shortestpossible time in which to complete your project You can do this by inspec-

tion of the logic diagram.

Enter the durations on to your Post-it notes in the logic diagram for eachkey stage Begin at the START note and trace each possible route or paththrough the diagram to the FINISH note, adding the durations of all thekey stages in the path The path that has the highest number – that is,

the longest duration – is the critical path of your project and represents the

shortest time in which it is possible to complete the project All other paths

are shorter All the key stages on the critical path must, by definition, finish

on time or the project schedule will slip

For example: referring to the previous logic diagram (Figure 7.6), theavailable paths are:

1 Start – planning – design phase 1 – purchase orders – install – testing –finish: 37 days

2 Start – planning – design phase 1 – design phase 2 – install – testing –finish: 42 days

3 Start – planning – design phase 1 – train staff – testing – finish: 36 days

4 Start – client survey – design phase 1 – purchase orders – install –testing – finish: 56 days

5 Start – client survey – design phase 1 – design phase 2 – install – testing– finish: 61 days

6 Start – client survey – design phase 1 – train staff – testing – finish: 46days

7 Start – client survey – training design – train staff – testing – finish: 42days

So, the critical path is number 5 in the list of available paths.

This is where reality hits you – is the project total time what yourcustomer actually requires? If it is a long way out, do not worry yet, asmost project managers expect this to happen Remember, your estimatesare based on people’s perceptions Your job is to attempt to compress thetime to a schedule that, on the one hand, is real and achievable and, on theother, satisfies your customer To do this you need to make use of another

valuable tool of project management: Programme Review and Evaluation

Technique (PERT) This tool allows you to analyse the logic diagram to

confirm:

• the critical path – confirmation of your inspection;

• the start and finish times of all the key stages;

• the amount of ‘spare time’ available in the non-critical key stages.All these data are very useful to you for optimizing the project scheduleand, more importantly, for the control of the project work once this starts

THE PROGRAMME EVALUATION AND REVIEW TECHNIQUE

The PERT method of critical path planning and scheduling is the mostcommonly used technique for project management control It is based onrepresenting the activities in a project by boxes (or nodes) that containessential information calculated about the project The interdependenciesbetween the activities are represented by arrows to show the flow of theproject through its various paths in the logic diagram The PERT diagram(sometimes referred to as a network) is identical to the logic diagram youderived earlier, each Post-it note for a key stage representing a node.The conventional data stored in the node box are as shown in Figure 7.7

The four corners of the node box are used to store the four characteristic

planning

5 days

design phase 1

9 days

purchase days

design phase 2

design training

design training

5 days

Figure 7.6 The logic diagram with durations inserted Note: durations are in

consistent units

times for the key stage These are calculated times using the durations

derived in estimating; remember to keep all durations in the same units

The lower middle box contains the total float for the key stage This is the

spare time in the key stage that allows you to take decisions about theactual start time or extending the duration within limits

In Figure 7.7 the earliest start time is day 12 and the latest start time isday 16 This gives an option to start the activity any time between day 12and day 16 The four-day difference is the spare time associated with theactivity Starting anywhere in this time zone will not affect the total projecttime provided the activity is fully completed by the latest finish time ofday 26 Say you were to start the activity on day 12 as the earliest planneddate If the time taken becomes extended from 10 days to 14 days, you use

up all the spare time but maintain the total project time If, however, theduration becomes extended to 16 days – two days more than the availablespare time – you will extend the total project time by two days

Obviously, this process applies to every key stage and lower-level ity in the WBS If every key stage takes longer than the available sparetime, the project will be very late If the spare time is calculated as zero

activ-then that key stage is termed critical and is one of those on the critical path.

The PERT technique is founded on calculating this information to permityou to take such decisions and control the project

The default or normal relationship used is FINISH-to-START (see Figure7.8) Under certain circumstances it is valid to impose constraints with theSTART-to-START or FINISH-to-FINISH relationships between activities

26 4 16

Activity description

Earliest finish time (EFT)

Latest finish time (LFT)

Latest start time

Activity description

Earliest finish time (EFT)

Latest finish time (LFT)

Latest start time

You can impose a forced delay using a LAG between the START or FINISH

of a predecessor activity and the START or FINISH of one or more sor activities

succes-The forced start or LEAD is used to start a SUCCESSOR ACTIVITYbefore the PREDECESSOR ACTIVITY is completed

Lags and leads should be used with care; it is easy to become confusedand introduce errors Split an activity instead of using leads to keep thediagram relatively easy to read and understand

Figure 7.8 Relationships in the PERT diagram

AE

AB

AD

‘AA’ IS THE PREDECESSOR

ACTIVITY OF ‘AB’

AND ‘AC’

‘AD’ IS THE SUCCESSOR ACTIVITY OF ‘AB’ AND ‘AC’

F S

AF

F S

AE

AB

AD

‘AA’ IS THE PREDECESSOR

ACTIVITY OF ‘AB’

AND ‘AC’

‘AD’ IS THE SUCCESSOR ACTIVITY OF ‘AB’ AND ‘AC’

F S

AF

F S

ACTIVITY OF ‘AB’

AND ‘AC’

‘AA’ IS THE PREDECESSOR

ACTIVITY OF ‘AB’

AND ‘AC’

‘AD’ IS THE SUCCESSOR ACTIVITY OF ‘AB’ AND ‘AC’

F S

AF

T

’

AY