The Teal Book: Part E
The purpose of controlling the work is to ensure that the outcomes are delivered and benefits realised within the constraints defined in the portfolio plan (for portfolios), business case (for programmes and projects) or work plan (for a work package).
Work needs to be controlled; otherwise, the sponsoring body and, where relevant, the delivery body, the portfolio director and senior responsible owner cannot have confidence that their objectives are likely to be met, in terms of outcomes delivered and benefits realised.
Controlling work ensures:
Control needs not only to be exercised in practice, it also needs to be exercised as part of accountability for use of public funding. Managing public money (requires sign in) states that ‘acting within the authority of the minister(s) to whom they are responsible, the accounting officer shall ensure that the organisation, and any arm’s length bodies it sponsors, operates effectively and to a high standard of probity’. This includes the management of an organisation’s portfolios, programmes and projects.
In project delivery ‘effective operation and probity’ includes complying with the functional standards and developing and using a compliant governance and management framework. ‘Controlling’ is central to making sure that work is progressing as planned and that problems are addressed without undue delay. Defined working approaches, coupled with good control, are an essential aspect of assurance (see Chapter 4: Governance and management).
The Project delivery glossary defines control as:
Any action taken by management, the board and other parties to manage risk and increase the likelihood that established objectives and goals will be achieved.
Controlling is about taking preventative and corrective action to rectify identified variances to the plan and deviations from expected behaviour, and to respond to changes in the context of the work. Good control relies on the effective tracking of progress and monitoring outside influences in the wider organisation and beyond. Controlling is the activity where information produced in every practice in Part E: Planning and control and Part F: Solution delivery is collated, assessed and analysed at every level in the project delivery work breakdown hierarchy (see Figure E.2 Primary interactions between planning and control practices in the context of management and solution delivery practices).
Control should also be forward-looking, with the aim of prompting action to keep the work on course to meet the defined objectives, addressing issues, taking advantage of opportunities and reducing the impact of threats, whilst verifying what has already happened.
Central to controlling is the concept of tolerance. The Project delivery glossary defines tolerance as:
The agreed deviation above and below a plan’s baseline without escalation to the next level of management.
Figure 17.1 shows the control cycle which can be applied at every level in the project delivery hierarchy and to every practice in The Teal Book. Work is undertaken in accordance with the plan using the delivery approaches defined in the governance and management framework. Progress is tracked using appropriate measures. The progress information is assessed and analysed and, if appropriate, collated into reports which may be periodic or by exception. Action is then taken to address any variances or problems which have been identified.
The frequency of the cycle should be both informal, with managers reacting to day-to-day circumstances, and also formal, for example synchronised with the reporting cycle. The further up the project delivery hierarchy, the more holistic and wide ranging the analysis should be. For example, when used at programme or portfolio level the additional perspectives needed can include:
The portfolio director, for a portfolio, and the senior responsible owner, for a programme or project, is accountable for ensuring an effective control framework is in place and being used.
The portfolio manager has accountability for controlling work in a portfolio and the programme or project manager for a programme or project, respectively. A work package manager controls the work in their assigned work package. Controlling is a fundamental aspect of those roles (for descriptions of the roles, see Chapter 11: The governance and management of portfolios and Chapter 13: The governance and management of programmes and projects).
A single person cannot control everything that needs controlling except on the smallest work item. The work therefore often needs dividing up between work package managers and members of a support office in order to have adequate coverage.
Each level of the project delivery hierarchy should be controlled by its controlling manager: for example, a project manager controls a project. They are accountable to the next higher-level controlling manager, such as a programme manager, who oversees the lower level controlling manager, and controls their own work. Similarly, that higher-level controlling manager is accountable to the next higher-level controlling manager and so on, up the project delivery hierarchy.
Not all activities need the same level of control as, even with a large team, there is often not enough time or management resource to watch everything all the time, and it is not an efficient use of resources. It is therefore important to focus on those activities which are most important, and/or which present the greatest threat to achieving the objectives, such as:
On a practical level this often means focusing on:
The focus can change as the work moves through its phases or when identified risks need watching with greater vigilance.
Having an achievable and logical plan is fundamental to effective control (see Figure 17.1). If the plan does not make sense in terms of logic and realism, controlling the work can be impossible. It is also important that the plan is baselined, otherwise terms such as late, early or over budget have no meaning. This does not necessarily mean the plan has to be highly detailed. The plan only needs to be detailed enough for measures of achievement to have real meaning.
Team members can be drawn from many business areas, from the contingent work force or from suppliers. They might have different ways of working in their ‘home’ business areas. They might also use different terminology. Whilst having a good governance and management framework and induction on using it helps, the level of experience of team members can vary. The practices in The Teal Book are in separate chapters, but in day-to-day work they are often inseparable; and information does not always come with a label attached, especially if people use different terms. For example, an issue might not be introduced as an ‘issue’ in a conversation, email or report; it might be a sentence in an email, a line in a spreadsheet or a verbal comment made in passing. It is often up to the manager to recognise and synthesise the information, make sense of it and treat it accordingly.
Progress information should not only be looked at solely in relation to its ‘date’ but also in relation to trends, including in the supply chain. This is particularly important in complex programmes and projects where multiple suppliers are involved. Is progress accelerating or slowing? Are costs escalating? Are the attitudes of stakeholders swinging away or becoming more positive? Such trends can prompt risks or issues to be raised and action taken. Similarly, the data might have no pattern and yet be outside expected or acceptable limits. This can represent instability, which again needs addressing. In the early phases, one of the more important indicators is how stable the user needs and requirements are. Regardless of the delivery approach taken, user needs and requirements need to have reached a degree of stability before follow-on activities can be started with confidence; for example, when producing a high-level design or signing a contract.
The choice of frequency for the formal control cycle is often tied to the reporting cycle (see Chapter 18: Reporting), but it does not have to be. For example, a report from one level of the project delivery hierarchy to another could be monthly, but the manager could hold weekly progress meetings with the team. The frequency can be different for every component in the hierarchy, with faster-moving or riskier work needing a shorter control cycle. Some delivery approaches define the frequency; for example, the Scrum method has a daily frequency in the form of a ‘daily stand-up’. Generally, the further down the hierarchy, the more frequent the control cycle. At a limit, if information on work is captured as it happens and available ‘live’, such as in finance systems, the requisite information (or part of it) can be drawn on as and when needed. It can also be continuously analysed using automated routines, such as in some software testing approaches. Greater use of reliable artificial intelligence applications is likely to prompt faster, real-time analysis of available data and lead to the use of the control cycle on a continuous basis, with exceptions being flagged when relevant.
The assessment and analysis of the work can range from very simple to extensive (as in formal evaluations, see Chapter 2: Policy and evaluation). The methods used need to reflect the work being done as well as why the information is needed. The overall aim, however, is to choose methods that highlight variances from plans and deviations from expected behaviour. Presentation is therefore important as poorly analysed and presented information can hide real problems. For example, if costs are less than planned for a particular date it doesn’t necessarily mean the work is behind schedule. Figure 17.3 shows an example of a method known as earned value management, combining views of schedule, cost and scope to show that while an underspend is partially due to work being late, the work is also costing less than planned. If the work is going exactly to plan, the planned, actual and earned value lines would coincide.
Graphical presentation, such as in Figure 17.3 is a good way to show trends. Figure 17.4 shows more examples relating to schedule. The burn-down chart is often used, notably in software development to track how many function points are left to be completed as opposed to the number which were planned to have been completed at a point in time. The units can be for any measure, such as miles of fibre rollout, volume of concrete, percentage of activities digitised or user satisfaction. The Gantt chart shows milestones and activities plotted against time. In this case a comparison between the planned dates and actual/forecast dates is immediately apparent. This example shows the gates and stage in the reference project life cycle but can be used at any level of detail in the project delivery work breakdown hierarchy, from portfolio to activity within a work package.
Further information on methods to support assessment and analysis is provided in guidance for infrastructure projects on earned value management and project control.
Check. Does the information presented make sense? Has it changed since the last report? When was it actually updated? Don’t just rely on the formal information channels. Talk to people, listen to people. Gauge whether what they say really reflects what is being reported. Look at physical progress to see if it reflects the information provided. Be visible and available for people to talk to informally; don’t wait for them to come to you.
Before starting to control the work, it is important to define the approach to be used. This should match the nature, scale and complexity of the portfolio, programme or project, and the expected delivery approach.
Controlling in government should also take account of central government and organisational requirements, particularly in relation to spending reviews, annual business planning and department specific HM Treasury delegated limits and with the wider approach to functional controls set out in the department’s memorandum of understanding with HM Treasury. The Green Book[/gpd_link] and its supporting guidance set out requirements for investment appraisal and business case development which apply to all programmes and projects involving central government expenditure.
In defining the approach to controlling, it is also important to consider how control is organised: what activities are needed and when, who carries them out, and how the outputs of these activities and other control products are to be structured, managed and stored (see 17.6.2 on preparing to control the work for more detail).
Not everything needs to be fully controlled, all the time. Decide, based on the current life cycle phase, type of activities being worked on and the prevailing risks, what needs to be controlled and therefore the information that needs to be gathered. This can change through the life cycle. For example, in a programme or project requirements are likely to be more prominent at the start but activities in later phases, such as data migration or embedding organisational changes, would not need monitoring until later. The control information should result directly from doing the work wherever possible, as this improves accuracy and reduces additional work for the team. Do not collect information on the basis that it might be useful; focus on information that is known to be useful.
Information provided by work package managers needs to be collated at project level by the project manager and passed up the project delivery hierarchy to programme, portfolio and up to organisational levels. It is therefore necessary to understand not only the issues facing those providing the information needed for effective control at one level in the hierarchy, but also what information the managers need at the higher levels.
Having understood who needs information and why, choose the appropriate units for measuring the progress and status of the work. Make sure the units are explicit. For example, costs measured as cashflow are different to those measured for statutory financial reporting. The units should relate directly to the work where possible. However, proxy measures might need to be used if direct measures cost too much, take too long to collect, or are burdensome. Timely information is necessary if prompt action is to be taken. Tracking measures should be the same as used for planning to ensure they are comparable. Take care not to choose measures to micro-manage the lower-level managers; this can be demotivating and lead to management burn-out.
Having understood what needs controlling, determine how frequently the progress and status information needs to be collected for analysis. Continuous? Daily? Weekly? This can change through the life cycle of the work as well as by component in the project delivery hierarchy (see 17.6.1.5 on choosing an appropriate control cycle and Figure 17.2). For portfolios, the information needed, and frequency is likely to be more stable as it is drawing on information from its work components. It is also more likely to be based on financial years rather than the life cycle, as in the case of a programme or project.
It is important to spot variances from the plan and deviations from expected performance early so that corrective or preventative action can be taken. Choose the way progress and status information is presented which makes it easy to identify the problem areas. Often the same presentation can be used for both monitoring and reporting, although monitoring data can sometimes need simplifying for reporting purposes.
The activities for controlling are summarised in Figure 17.5. These are iterative, with similar activities repeated through the project delivery hierarchy, and can be applied to every practice in The Teal Book.
The approach to controlling the work should be defined including any processes, methods, tools and techniques to be used. This forms part of the overall governance and management framework for the work (see Chapter 4: Governance and management). The important aspects of this activity are discussed in section discussed in section 17.6.2 on preparing to control the work.
The framework should be maintained to address relevant feedback from its use. It is not important whether the higher-level controlling manager or the controlling manger develops the framework as long as the arrangements are appropriate and proportionate to the work being undertaken.
The higher-level controlling manager should ensure they are receiving the information they need, whether informally, as reports or by drawing them off management information systems when needed. They are likely to be dealing with information relating to a large number of work components and should make sure that the information is consistent and usable. They should look for overall trends both with individual work components but also across work components,
It is essential to maintain an overall view of how effective the control activities are at reflecting the progress being made on the work.
Work should be undertaken in accordance with the plans and using the delivery approaches and controls defined in the governance and management framework.
The progress of the work should be tracked using the measures and frequencies prescribed in the control framework, ensuring that the information is timely and reflects observable progress where physically apparent. Information should be drawn on from each relevant Part E: planning and control and PArt F: solution delivery practice and verified to ensure it is sufficient for a reliable view to be taken regarding the current status of the work.
The status of the work should be assessed and analysed, using the defined approaches. Material variances and deviations, and inconsistencies in information should be noted together with understanding the reasons for them, for example using root cause analysis techniques. Update the risk and issues register, as necessary. This information can be used to populate formal reports (see Chapter 18: Reporting). The Magenta Book (requires sign in), the Government Functional Standard for Analysis and the Aqua Book (requires sign in) include or refer to, a range of analysis methods for different situations.
After gaining an understanding of the situation, decide what action, if any, needs to be taken. This can include responding to a risk or issue or submitting a change request. Informal actions can also be taken and noted, say on an action list for the work component.
Once the control framework is no longer needed it should be closed. For a portfolio, this is unlikely unless the portfolio ceases to service its purpose. Closure of the framework can be progressive throughout a programme or project as each work component comes to an end.
Page permissions updated for public launch.
First published for closed beta consultation.