{nvernational Journal of Project Management Nol. 17. No.1. pp. $5-59, 1999 (1998 Elsevier Seience Ltd and IPMA. All rights reserved Printed in Great Britain (0263-7863/98 $19.00 + 0.00 $0263-7863(98)0000S-2 Company-wide project management: the planning and control of programmes of projects of different type John H Payne Line Management Group Ltd, Stebenheath House, 82 West India Dock Road, London, El4, UK J Rodney Turner* Department of Organisation and Business, Faculty of Economie Sciences, Erasmus University, Room HIS- 28, Burgmeister Oudlaan, Rotterdam, The Netherlands It has been perceived wisdom that where an organisation is undertaking a portfolio of projects, they should use a common approach to the management of all projects in the programme. Pre- sumed benefits include comparable progress reporting, and consistent calculation of resource requirements enabling sharing of resources. People can also move freely between projects with- out having to learn a new management approach. However, research undertaken by the authors show that people more often report better results for their projects when they tailor the pro- cedures to the type of project they are working on, matching the procedures to the size of the project, or the type of resource working on the project. In this paper, the authors report their findings and give an explanation of why, on many projects, it may be better to tailor procedures. Since it is still worthwhile to obtain some consistency of project management approach to achieve the benefits above, the authors suggest how to adopt a consistent approach at the stra- tegic level, while tailoring the procedures at the tactical or detail level. They give an example of the use of approach on the planting and control of proj ct from their own experience. © 1998 Elsevier Science Ltd and IPMA. All rights reserved Introduction It has previously been suggested that an organisation undertaking several projects should adopt a common project management approach for all projects in the programme, regardless of the type of project, its size, or the type of resource used.' Advantages are said to be: # a consistent reporting mechanism can be adopted to give comparable progress reports across all projects, ina programme; © resource requirements can be calculated on a con- sistent basis, facilitating the management of capacity constraints; © people can move between projects without having to relearn the management approach used project by project; *Conrespondence address: Wiklwood, Manor Close, East Horsley, Leatherhead. Surey. KT24 6S. UK. © small projects can be used as a training ground for future managers of large projects. ‘An inherent, though often unrecognised assumption behind this view is that the projects within the pro- gramme are fundamentally homogeneous. Turner? implied this in his review of several case_ studies However, Payne’ in a review of the state of the art of Programme Management identified that inhomogeneity adds complexity to Programme Management, for instance where the projects within the programme are of differing size, urgency and skill mix. As part of a larger research programme, Payne developed a ques- tionnaire to test the assumption that even in these cir- cumstances it is preferable to adopt a common approach to the management of all projects within a programme, Surprisingly he found the exact opposite that where people used consistent procedures regard- less of project type, size and skill type, they reported less success than where people tailored their pro- cedures, 38Company-wide project management: John H. Payne and J. Rodney Turner In this paper, we report on Payne's findings, provide an explanation based on Turner and Cochrane’s Goals and Methods Matrix" show how programme man- agers can still achieve the advantages listed above by using a consistent approach to project strategy while allowing diversity at the tactical level, and illustrate through the planning and control of ‘a project. pre- viously described by Turner.* ‘The findings To test the hypothesis that it is better to use a com- ‘mon set of procedures on all projects in a programme, Payne designed a questionnaire, sent to almost 5000 project managers, including all the members of the UK's Association for Project Management. He received 150 replies. Respondents were asked to describe the range of project types in their organisation. They were given guidance to categorise projects as major, large, med jum and small. Respondents were asked to say what they considered to be a large project. A major project was defined as one ten times larger, a medium project fone (en times smaller, and a small project one ten times smaller again. (It was suggested to the respon- dents that a major project was one of value roughly equal to the capitalisation of the parent organisation, a large project one tenth the capitalisation, etc.) Major projects have such a large effect on the organisation that even expected outcomes distort normal measures of organisational performance. They were therefore ignored in the analysis. Respondents were also asked to describe the range of skill types used on their pro- jects. engineering, building, information systems, per- sonnel, ete, Respondents were then asked to describe the project management procedures (computer or paper-based), used to manage projects in their organisation, and in particular to say whether the same approach is used ‘on all projects regardless of size or skill mix, or whether the approach is tailored to the type of project. They were also asked to report how often their pro- jects were successful or unsuccessful. (It was allowed that project might be of mixed success, that is they may achieve some but not all of their objectives Hence the sum of successful and unsuccessful projects is less than 100%.) Table I reports the findings. You will see that the respondents reported, on average, greater success if they tailored their procedures to the type of project and increased failure, if they used common procedures for all projects. Unfortunately, insufficient replies were recived to make all the results statistically significant and so although there was an apparent trend, it was not proved. The results for large projects are signifi- cant to 95% confidence levels but not 99%, the results for small projects are significant to 99% confidence levels, but the results for medium-sized projects and for different resource types are not significant even to 95% confidence levels. Indeed for medium sized pro- jects there is almost no difference at all in the results for success or failure. Perhaps this means that where common procedures are used, they are designed for medium sized projects, yet it is almost certainly better on small projects to use small project procedures, and probably better on large projects to use large project 56 Table 1. Project success and fail «ype with tallored procedures a Projects by site Large Medium ‘Small Differing resource types procedures. What was also proved is that it is defi- nitely not universally better to use common, company- wide procedures. Hence, from these results we conclude that where organisations are undertaking programmes of projects in which the projects have different sizes and involve different types of resources: 1. tailoring the project management procedures to the size of project and to the resource type does not reduce the chance of success, and probably increases it 2. applying common procedures across projects of all size, and across all resource types increases the risk of failure ‘The explanation Projects by size We probably all understand why it i lor the procedures by size of project. 1. In the management of small to medium sized pro- jects, the main emphasis is on the prioritisation of resources across several projects. Small projects also cannot stand the bureaucracy of procedures designed for larger, more complex projects 2. In the management of large projects, the main emphasis is on the coordination of ‘a complex sequence of activities, balancing resources across the activities, but within the control of the project manager, to enable the critical activities to take place in time, and to stop the bulk work becoming resource constrained, Large projects have much greater data management requirements than small to medium sized projects. 3. In the management of major projects, the emphasis is on coordinating the activities of people across several sub-projects, and on managing the consider- able risk (a failure of the project will sink the parent organisation). necessary to tai- Projects by resource type Turner’ dealt with how the matrix developed by him- self and Cochrane* implies that different types of pro- ject require a different approach to their planning and control, Figure 1 1. Engineering projects are labelled Type 1 projects, and with well defined goals and methods of achiev- ing those goals lend themselves to activity-basedCompany-wide project management: John H. Payne and J. Rodney Turner "Type? Project Types Prajeae Proust Research & No Development Methods Well War Defined Yes ‘Type Projets No Goals Well Defined Figure 1. Turner and Cochrane's Goals and Methods Matrix approaches to planning. It is these types of projects that many of the traditional books on project man- agement have been written about, that many of the traditional software products, such as Artemis, have been developed for, and which have a long history of proceduralisation in the engineering construction and building industries Product development projects are Type 2. projects. The goals are well understood, but identifying the method of achieving the goals is the main point of the project. Examples are weapons systems develop- ment and projects from the electronics and manu- facturing industries. The early project management procedures developed in the 1950s by the US mili- tary were aimed at these types, and more recently the goal directed approaches.* The plans for this type of project are best based on a Bill of Materials (Product Break-down Structure), based on the Known goals; that is a milestone based approach to planning, where the milestones represent com- ponents of the eventual product. 3, Information Systems projects are labelled Type 3 projects, With the goals poorly defined, the plan- ning approaches tend to be based around the pro- ject life-cycle; that is a milestone-based approach to planning is adopted, but the milestones now rep- resent completion of life-cycle stages. Methodologies such as Prompt, Prince and Prince 2 are aimed at this type of project, as are computer systems such as PMW. 4. Type 4 projects tend to be managed as Type 2 or Type 3 projects depending on their nature. Research projects tend to be managed through the life-cycle and the achievement of go/no-go de- cisions, whereas organisational change projects tend to be managed through a Bill of Materials or pro- duct-based milestone plan, If you try to adopt an activity-based approach to the planning and control Type 2, 3 or 4 projects, then it will increase the likelihood of failure. Thus we see for projects of different sizes, and for projects of different resource types, we must tailor our planning and control procedures to meet the needs of, the individual project types. Similar arguments apply to other project management functions and processes. The solution In the introduction, we listed some presumed benefits for adopting a common approach to the management Of all the projects in a programme. These potential benefits still remain, So, how can we can we resolve the dilemma of achieving those benefits while still developing a system that meets the needs of the indi- vidual project types? The answer is to develop a stra- tegic plan for every project based on the common approach, but allow different projects to adopt differ- ent approaches at the detail or tactical level. This means at the three fundamental levels of planning,” the following plans are developed: 1, Integrative Level: A Project Definition Report® will be developed for all projects, based on a common model. This will ensure that all projects are defined in a consistent way, giving a common basis for comparison and prioritisation. For larger projects the report will be more detailed than it is for smal- ler projects. 2, Strategic Level: A Milestone Plan and Project Responsibility Chart** will be developed for all projects. For Types I and 2 projects, the milestones will represent components of the eventual product, for Types 3 and 4 projects, the completion of life- cycle stages. This will give a consistent approach for assigning resources and responsibilities, and for tracking and comparing progress. The programme resource plan will be developed at the milestone level. For larger projects, the project may be broken into several sub-projects or stages, each with a mile- stone plan. For smaller projects, one plan will cover the entire project, 3. Tactical Level: At this level and below, project plan- ning methods will be chosen based on the nature of the project. ‘© For small projects, there may be no further levels of planning. © For large projects there may be one or more levels of planning. © For engineering construction and building pro- jects, the lower levels will be developed in some detail at an early stage, based on the known ac- tivities to be performed. 37