This week we look at the methods of estimating project times and costs. The initial cost estimates are only intended to determine whether or not the project will proceed. Detailed costings come later and only if project is going ahead. In addition, this week looks at the establishment of a project network in order to take advantage of every opportunity to make the project efficient. This includes the process of network computation, considering the practical aspects for the analysis of the network and look at extended techniques relating to project networks.
Project estimation will always be better if you surround yourself with the right people to provide quality information. Allowing for risk is also important so there is a degree of buffering in the project that you may not have accounted for. When projects are based on estimations, the stakeholders will be aware of the bases of the project so it is essential to keep them informed along the way.
Estimating Project Times and Costs
Poor estimates can often be the reason behind project failure. Larson and Gray (2020) set out seven factors that can have an impact on devising estimates for a project.
- Responsibility – estimates should be made by people that understand the task to be completed. They should have the experience to add expertise to the estimation.
- Using several people to estimate – Estimates completed by a team reaching consensus usually have a far better chance of being more accurate.
- Normal conditions – estimates should be based around what could be expected to be normal conditions, not optimum or worst case scenario. That qualifier should be included in the estimates.
- Time units – what type of time unit will be used for the project. Will it be based on hours, days, weeks, 24/7 shifts? The text suggests that where time is critical such as a transplant operation, that minutes should be considered as the unit of time.
- Independence – estimations should be treated on the basis that they are independent from each other even if they may be integrated in the Work Breakdown Schedule (WBS).
- Contingencies – estimates are not the place to include contingencies. Any estimate should be based on normal conditions. Contingencies will be allowed for separately.
- Risk assessment – should be included with estimates to apprise stakeholders around the risk associated with the task. Obviously, some tasks will have more risk associated with them in terms of time and cost.
Quality of Estimates
Larson and Gray (2020 p. 137) state that a “typical statement in the field is the desire to have a 95% probability of meeting time and cost estimates”. Past knowledge and experience, the right people and other factors can impact upon the quality of estimates. These can include:-
- Planning Horizon – estimates done on the basis that a project will be done immediately have a far greater chance of being accurate than one to be done a year hence.
- Project Complexity – the implementation of new technology often has the problem of time expansion which in turn can impact on costs.
- People – as stated above, the quality of the people and their experience can impact upon the quality of the estimates.
- Project Structure and Organisation – how the project team is structured can have an impact on time and cost. Dedicated project teams can provide speed whereas a team who are doing a project as a part of their everyday job will have competing priorities.
- Organisational culture – some organisations tolerate and even encourage padding of estimates whereas others expect accuracy.
- Other factors – there are external factors that can come into play at any time that can impact upon project estimates, in particular those of an external nature. The outbreak of Covid is one such example.
Essentially, there are three types of estimating methodology. These are top-down, bottom-up or a hybrid of the two.
Figure 1: Conditions for Preferring Top-Down or Bottom-Up Time and Cost Estimates
Top Down Estimating
The top down approach is often used to obtain an overview of the project to evaluate its viability. A lot of the detailed information will not be available and the method can be used until the tasks are detailed in the WBS. The application of a consensus method would call upon the expertise of senior people. A further method, the Delphi model, obtains opinions from experts in the field of the project, but still use a consensus model with other experts. This method is useful when the project is quite complex.
Another alternative is the ratio method often used by the construction industry (Larson & Gray 2020). If there is a known area and known estimates based on that area, then the two can be multiplied out to provide an estimate. An example would be where building costs are $500 a square metre and a building is 1,000 square metres, then the estimate would be $500,000. These same figures, based on experience, can also provide an estimate of the time it will take to complete the project.
The apportion method considers the expertise of managers on how costs are usually spread over different aspects of a project. Many projects exist where known percentages can be applied to envisioned projects that are similar or the same.
The final top-down approach is the function point method used in the software industry. The process is to assess the number of inputs, outputs, inquiries and data files combined with the number of interfaces and then applying a weighting to each of these combined with a factor of difficulty. This can then assist in determining what the labour cost might be.
Bottom Up Estimating
The Template Method can be used where the project is similar to past projects. Initially, this would involve a comparison of the original project to the envisaged project to identify differences. Using a database for this method can assist in reducing mistakes in estimations.
The Parametric Method has similarities to the ratio method in the top-down approach but in this case ratios are applied based on the experience of the people involved. The ratios are applied more in relation to specific work in the project as opposed to the project itself.
Range Estimating is used when tasks involve considerable degrees of uncertainty around time and cost. Experienced people have a better idea of the amount of time or money involved. When range estimating is used, the text suggests using a high, low and average amount for determining the estimates (Larson & Gray 2020).
Figure 2: Top Down and Bottom Up Estimates
Phase or Hybrid Estimating
With this approach to estimating, it will commence by using a top-down approach with subsequent refining of the estimates as different phases of the project are initiated. Some projects simply resist estimates at all levels because of the complexity of the project. This is when phase estimating is used due to the impracticality of trying to estimate time and costs which will not be real. So, there is a proper estimate for the immediate phase of the project and an overall estimate for the balance.
Developing and Constructing the Project Network
With estimations of time in the WBS completed, a project network diagram can now be completed. This diagram is simply a graphical representation of the interaction and linking between the various tasks in the WBS. In turn, this provides the Project Manager to schedule the tasks for completion in the most efficient manner. The text suggests that it would be virtually impossible to gain an overview of how tasks are linked without a visual representation (Larson & Gray 2020).
Of course, not all projects will require such detail. Small projects may be served better by a simple breakdown of the tasks in the order in which completion is required.
The text does refer however to identifying the critical path which is the longest time period to complete a list of tasks. Any delay in these tasks will affect the entire project by the same amount of time.
The WBS provides the information for the project network as it lists all the tasks in a project. The project network then is effectively a visual representation of the WBS.
The design of the Project Network centres around the concept of Activity On Node (AON). Nodes are generally drawn on diagrams as rectangles and linked by arrows showing the dependencies of each activity. It is important to note that when it comes to the project that tasks are activities, but in the diagram, activities in the diagram may not necessarily be tasks. Activities in the network diagram will have identification and details.
Figure 3: Activity on the Node Network
Network Computation Status
This involves determining the start and finish times of each of the activities on the project network. Estimations of time are applied from the WBS and added to the network. The project manager can now, with the use of some computations, implement a process known as the forward and backward pass. A forward pass determines the earliest time a a task or activity can commence and finish. The backward pass determines the opposite, the lastest time an task or activity can commence and finish. The difference between the two times becomes the float or slack but some tasks don’t have this luxury. These tasks therefore become ones of a critical nature.
Practical considerations in network analysis
There are additional practical considerations that need to be borne in mind when analysing networks. The network logic errors can become quite common due to the complexity of a project. The more complex, the greater risk of them appearing. When engaging in activity numbering it is vital to ensure that all numbers are unique. Gantt charts can be used when a project is small and avoids the need for complex project network diagrams. Calendar dates need to ensure allowances are made for the number of days taken for the project and must make allowances for non-work days.
Computer models will often have a requirement that a node identifies the the common start and finish dates for a project. If it isn’t required, it is still sound practice to do this in order to avoid what are known as “dangler paths” where it may appear a project has no defined start or finish date. In addition, this can also be true when a project completes with more than one activity as the paths will not be connected. Everything should link back to a common start and finish node.
Extended Network Techniques
Projects often rely on the concept that a series of activities need to be completed prior to the next activity being commenced. This is not always the case though as some activities can be partially completed and then next activity can start where the previous one started. This practice can result in improving time components of the project if identified. The text refers to this practice as laddering but also recognises that this can be too restrictive for some projects.
Lags are another consideration in project planning and networks. These are used to identify the lag time between the completion of one activity without the subsequent action occurring. For example, if an activity requires the placing of an order and then there is a lag of 15 days before receipt of the order, this becomes the lag rather than identifying that the task takes 16 days to complete.
Project Management Software
Project management for small projects can be done on paper, but for large complex projects, Project Management Software Tools (PMST) will be a far more productive solution. There is quite an array of tools on the market but all should include features such as the ability to schedule projects, allocate resources and costs, detail the budget and actuals and manage all documentation. One package that is free for Project Management is ProjectLibre.
The calculation of estimates, for both time and cost of a project are critical in its overall success. It is this phase that provides the basic information where future problems can be avoided. Based on what is prepared at this stage, if the project runs to schedule, then there will be no surprises that require explanation.
Additionally, this module has covered the graphical representation of a project by way of the project network. This is another vital document in taking in how a project will progress at a glance as opposed to trying to absorb pages and pages of notes.References
- Larson, EW & Gray, CF 2020, Project management: The managerial process, 8th edn, McGraw-Hill, New York.