Chapter 5: Lean Portfolio Management
Introduction to Lean Portfolio Management: The Role of Projects
The projects realize three types of initiatives that support the flow of value created by the organization:
- Strategic transformation initiatives
- Improvement initiatives
- Exploratory initiatives
Figure 8-1 shows the relationship between the lean project initiatives and the flow of value.
Transformation initiatives create new value streams or transform the whole value creation system. They originate directly from the organization's strategy.
Improvement initiatives stem from the value stream system. These are major and continuous improvements (kaikaku and kaizen) that may include projects for:
- developing new products and services within existing product/service families
- improving existing products, services, value streams and processes
- eliminating obsolete value streams
- ensuring compliance and responding to emergency situations
Exploratory initiatives acquire validated knowledge through rapid plan-do-check-act (PDCA) cycles and serve transformations and improvements. They are stand-alone projects or an Exploration phase of the Lean Project Life Cycle (as we discuss in The Lean Project Life Cycle chapter).
The organization's project initiatives differ from each other, and from ongoing initiatives in terms of impact scale, time horizon, frequency, and uncertainty. We show these characteristics in Figure 8-2.
Figure 8-2: Characteristics of Organization’s Initiatives
All these types of initiatives are important and they work as a system. For sustainable development, the organization must maintain an optimal dynamic balance between them (Apply systems thinking principle).
Figure 8-3: Lean Project Portfolio Allocation
The goal of Lean Project Portfolio Management is to maximize the net value of the organization's initiatives, thus helping to optimize the flow of net value that the organization creates. Portfolio initiatives realize their effect through operational value streams, and each initiative should improve the capacity of the value stream system to create value.
Lean Project Portfolio Management is a means for sustainable development and continuous improvement of the organizational value-creating system. It is fundamentally different from traditional project portfolio management, as shown in the table below.
Table 8-4: Traditional Portfolio Management vs. Lean Portfolio Management
Lean Project Portfolio Management has three processes that interact with each other and run simultaneously:
- Develop organizational strategy and set objectives
- Generate ideas, select and prioritize projects
- Fund, execute and evaluate
Develop Organizational Strategy and Set Objectives
The vision of the organization gives meaning to its existence, and the strategy is a high-level plan for realizing the vision. The strategy and its objectives provide a framework for defining value-stream objectives. In the lean organization, the strategy, the objectives and the performance measures result from a bidirectional process which is both top-down and bottom-up.
The objectives and Key Performance Indicators (KPIs) or Objectives and Key Results (OKRs) at both the strategic and value stream levels set the foundation for lean portfolio management.
Generate Ideas, Select and Prioritize Projects
Lean organizations constantly screen the environment and sense and seize opportunities. They encourage innovation and intrapreneurship. Kaizen and kaikaku activities create a continuous flow of improvement initiatives. Lean leaders initiate ambitious strategic transformations. Teams use stakeholder feedback to improve customer service and identify fresh opportunities. (See Idea Generation and Triage)
To be suitable for selection, potential projects must be:
Projects are aligned when they contribute to the achievement of the strategic objectives (for the transformation projects) or the objectives of the value streams (for all other projects) as measured by the relevant KPIs or OKRs.
We need not invent project objectives different from the strategic and value-stream objectives. An organizational objective may require several initiatives, and the individual projects will provide incremental contributions to achieving that objective. The opportunities for creativity and innovation remain endless, as there are many ways we can achieve each objective.
Alignment requires a binary decision. It's about yes or no, there are no intermediate options.
The rule is: Use the organization’s objectives for project objectives.
Feasible projects are those that are practically possible (at reasonable cost), in terms of the project environment, the nature of the project deliverables and the process of their creation, and in terms of the organization's capacity.
At the selection stage, viable projects are those whose potential benefits outweigh the potential costs.
The LeanPM Framework sets an additional requirement, which is to select Minimum Viable Projects.
Minimum Viable Project (MVP)
A Minimum Viable Project is a viable project that is minimized in terms of complexity, scope, effort, cycle time, dependencies, and risk (Simplify principle). The focus is not on minimizing but on viability and its maximization.
The questions we need to answer include:
- Is the estimated net value of the initiative sufficient to justify its inclusion in the project portfolio (above the threshold; a minimum desirable net value)?
- Can we implement the initiative more effectively and efficiently as an ongoing continuous improvement or exploration?
- Can we achieve the objectives through a better and simpler project approach?
- Can we achieve the project objectives through better and narrower deliverable configuration, better and simpler deliverables, and better and simpler deliverable creation processes?
- Can we minimize external project dependencies?
- Can we break down the initiative into discrete projects which would reduce overall complexity, uncertainty and cost, and would speed up value creation?
- Is the project independent (not a part of mutually exclusive projects)?
- Does the project overlap with other initiatives?
One aim of the Minimum Viable Project is to draw the line between projects and ongoing improvement initiatives. It should support a growing number of initiatives that become a flow of ongoing improvements, at the expense of initiatives formalized as projects.
The Minimum Viable Project reduces waste and risk, shortens cycle time, increases the probability of success, and limits the cost of failure. By minimizing individual viable projects, we maximize the net value of the project portfolio.
The concept of the Minimum Viable Project resembles that of the Minimum Viable Product, but there are significant differences between them. The Minimum Viable Project seeks to improve the project’s manageability. The Minimum Viable Product helps you quickly validate or invalidate hypotheses and fail fast.
LeanPM Framework uses the Minimum Viable Product and other similar concepts in the exploratory projects. In a Minimum Viable Exploratory Project, we create validated knowledge through rapid plan-do-check-act cycles, using increments that include:
- Minimum Viable Product
- Minimum Viable Service
- Minimum Viable Value Stream
- Minimum Viable Experiment
- Minimum Viable Research
- Minimum Viable Study
- Minimum Viable Test
The initial screening will create a pool of potentially aligned, feasible and viable projects, and then selections can be made for inclusion in the portfolio (see A3 Analysis and Pre-Selection/Selection).
The selection should be grounded on a project’s absolute and relative importance, or simply put, on the importance and priority.
Prioritization addresses the following issues:
- Allocation of resources to the transformation portfolio and the value streams portfolio
- Allocation by strategic themes within the transformation portfolio
- Allocation of resources to individual value stream and the cross-value stream portfolios as well as to major and continuous improvement projects, exploratory projects and ongoing continuous improvements
- Project sequencing (schedule priority)
Allocation is a bi-directional process with complete decentralization at the value stream level. The governance teams discuss and decide on the relative importance of the different portfolios and sub-portfolios and align the allocation with the capacity.
Prioritization seeks to achieve an optimum balanced mix of sequenced initiatives that will best deliver the strategy. The organization must continuously challenge and revisit priorities and allocation.
If we determine the importance of each project, we can sort the projects from highest to lowest in importance and select projects until we have exhausted the resources of the respective portfolio.
It’s a common practice to use scoring models to evaluate project importance. The scoring model is composed of multiple criteria with assigned weights and a scoring scale. The governance team assigns numerical scores to each criterion, and the total weighted score represents project value and importance.
For example, we could use a scoring model for triage and sorting of project ideas into one of the four categories of MoSCoW prioritization: Must have, Should have, Could have and Won't have.
However, the total project score is not a good measure of project value. When we use it, we encounter the following problems:
- The total project score does not carry objective information about the project’s value. The score does not represent absolute value, but (at best) relative value of one project compared to another.
- We measure the project score and project cost in different units, which makes it impossible to calculate the net value (the net value rather than the value shows project’s importance).
- We can manipulate the result through the scores and weights.
- It seems logical to use total project score to sequence projects (and we often do this), but, as we’ll see, this is incorrect.
- When we use the expected net monetary value together with other criteria in the scoring model, this means that we haven’t properly measured the expected benefits and costs. Adding a risk criterion, for example, shows that when estimating the net value, we didn’t take into account the parametric and model uncertainty (and if we did, we should not use a duplicate criterion).
- The arbitrary, indiscriminate use of diverging criteria and subjective scoring and weighting may result in a biased outcome.
Among the most commonly used criteria in scoring models are strategic alignment, return, risk and compliance. Organizations also use a variety of other criteria that ultimately affect benefits, costs, or risks.
But there are many questions that need to be asked:
- What's the point of assigning 0 score for “no alignment with strategic goals” and a score of 4 for “alignment with more than 3 strategic goals”?
- Can we accept “no alignment”?
- Can we compensate for non-alignment with a higher score for project size and technical feasibility, for example?
- Is the project impact proportional to the number of strategic objectives it’s aligned to?
- If a project contributes to the achievement of only one strategic objective, does this make it inconsistent with the other strategic objectives (and would a good strategic plan allow such inconsistency)?
Instead of a rational, we get a lot of questions. As noted above, alignment is a precondition for considering the project and should not be part of the appraisal model.
As we suggest, using the organization's objectives for project objectives has another important advantage. If we have previously evaluated how the progress in achieving objectives benefits the organization, this will help us assess the effect of the incremental contribution of the projects.
Let's look at three examples of value stream objectives:
- To improve Lead Conversion Ratio (the ratio of the number of leads to the number of those who turn into customers).
- To lower the Absenteeism Rate (the ratio of the number of absences to the number of workdays in a given period, expressed as a percentage).
- To improve Donor Retention Rate for a nonprofit organization (percentage of donors who have donated more than once).
Operational value streams should have evaluated (in monetary units) the effect of incremental change in the values of the indicators associated with these objectives. This would enable us to assess the impact of projects that, for example, aim to reduce the Absenteeism Rate from 4.2% to 3.3%, improve Lead Conversion Ratio from 20:1 to 15:1, and improve Donor Retention Rate from 38% to 43%.
Using compliance and urgency (emergency) in scoring models is also problematic. If we need a project to save the organization and continue operations, this project is mandatory and of utmost importance. We should include it in the portfolio. Where the threat is not apparent and imminent, the benefit-cost assessment will establish the project’s importance, as for any project. So, we have a better way to take compliance and urgency into account.
Risk-Adjusted Expected Net Value
Risk has no independent role in the importance of the project. Its role lies in the fact that it affects the benefits and costs. The same applies to many other criteria used in the scoring models (including time-related factors). We should therefore strive to take into account the impact of risk and other factors on benefits and costs, to work out the risk-adjusted expected net value of the project as the only measure of its importance.
Once we have selected projects for each portfolio using risk-adjusted expected net value, it’s very tempting to apply this metric of importance in order to establish schedule priorities and to sequence the projects. But that would be a mistake.
Projects Sequencing (Schedule Prioritization)
As our goal is to maximize the total net value of the portfolio for any time period, we need to consider its total Cost of Time. If we execute one project before another that has a higher Cost of Time, we may incur costs that we can avoid by changing the order of execution of the two projects.
Don Reinertsen gives three principles for prioritizing (sequencing) jobs: Shortest Job First (SJF), High Delay Cost First (HDCF) and Weighted Shortest Job First (WSJF) .
Shortest Job First (SJF) is the best sequencing strategy when all jobs have the same cost of delay. Starting with the longest jobs would delay the others more than when starting with the shortest jobs and would lead to higher overall delay cost.
High Delay Cost First (HDCF) should be applied when the costs of delay differ, but all jobs have the same duration. The jobs with higher cost of delay have a priority.
Weighted Shortest Job First (WSJF) is the best scheduling strategy when both durations and delay costs are different. WSJF is equal to delay cost divided by job duration. WSJF is also referred to as CD3 (Cost of Delay Divided by Duration). The logic is that the higher the delay cost and the shorter the duration of the job, the smaller its contribution to the total costs when we do it earlier.
Here is an example of applying the WSJF rule.
Cost of Delay ($/week)
From the information in the table, we can calculate which of the sequencing combinations produces the optimum results (here we present the calculations for four of the possible six combinations):
- Project B is delayed by 8 weeks and the CoD is $20K
- Project C is delayed by 20 weeks and the CoD is $64K
- Project A is delayed by 24 weeks and the CoD is $24K
The total CoD is $108K
- Project A is delayed by 4 weeks and the CoD is $4K
- Project B is delayed by 12 weeks and the CoD is $30K
- Project C is delayed by 24 weeks and the CoD is $76.8K
The total CoD is $110.8K
- Project C is delayed by 12 weeks and the CoD is $38.4K
- Project B is delayed by 20 weeks and the CoD is $50K
- Project A is delayed by 24 weeks and the CoD is $24K
The total CoD is $112.4K
- Project A is delayed by 4 weeks and the CoD is $4K
- Project C is delayed by 16 weeks and the CoD is $51.2K
- Project B is delayed by 24 weeks and the CoD is $60K
The total CoD is $115.2K
So, according to the WSJF rule, the best sequence is B-C-A because it has the lowest total CoD of $108K. The sequence of A-C-B has the highest CoD of $115.2K.
Let's add two more projects. The best sequence now is D-E-B-C-A and the sequence with the highest CoD is A-C-B-E-D.
Cost of Delay ($/week)
Sequencing projects does not imply that we need to wait for one project to finish before starting another. Depending on our resource capacity and the target average lead time, we will have a certain number of projects in process.
WSJF principle is a major improvement in decision-making. We can see that to apply it, two conditions must be met:
- The delay costs of the individual jobs must be constant for all time periods.
- All jobs must have the same reference point of delay (the present moment).
In fact, the WSJF, HDCF and SJF principles take into account the effect of postponing the jobs. We can formalize this effect as a Cost of Postponed Execution (CPE), which includes components such as lost opportunity cost, deferred use cost and customer dissatisfaction cost. As we discuss in the Cost of Time chapter, projects can have distinct points of reference for postponement and the postponement costs may vary over the time periods.
WSJF uses the expected job duration, which it takes as a constant. The delay in this case refers to the delay in starting and ending the work, and not to the extension of its cycle time (which we defined as the delay cost factor).
It’s a common situation to have unique values of the cost of postponed execution of a project for each future time period. Along with the varying reference points of postponement, this makes the sequencing a more complex task.
Let's look at an example of three projects that we need to arrange by order of execution. We need to find the best sequence among the six possible combinations.
Expected duration (number of time periods)
Ideal point in time to finish Just-in-Time
Estimated ideal point in time to start
Beginning of period 2
Beginning of period 4
Beginning of period 1
Cost of Time and WSJF
Since we can't directly apply the WSJF formula, we could calculate the average CoD/CPE values for the period 1-10, which gives us Alfa-Beta-Gama as the optimal sequencing combination.
It's obvious that in this case we cannot apply the WSJF and SJF rules correctly. When we perform the individual calculations for each combination (which we show below), we can see that the optimal sequence is Alfa-Gama-Beta.
We can conclude that when the postponed execution costs for the individual time periods, the durations, and/or the reference points for postponement are not all homogeneous, we can’t apply the SJF, HDCF and WSJF rules. We need to identify the correct priority by comparing the total costs of postponed execution of the individual combinations.
Note also that the CPE priorities will depend on the time horizon and the trends in the time-related costs. Varying the time horizon may lead to different optimal prioritization.
The portfolio allocation should not be fixed for the entire year. It needs to be regularly assessed and altered as needed, for example, following quarterly or monthly feedback loops. Within the funds allocated to each portfolio, the individual projects should be funded incrementally and each portion of the investment should be justified by the net profit it would generate.
Incremental financing reduces risk and waste. Short feedback loops on project progress help to improve budget estimates and to better assess the probability of achieving project objectives through rolling forecasts. The more accurate net project value estimates, and the changing environment and organizational objectives, may require a rearrangement of portfolio priorities. Therefore, project funding should be incremental, taking into account the shifting budgetary needs and priorities.
The key aspects of lean portfolio execution are ongoing project coordination and synchronization and portfolio optimization, by altering resource allocation and changing, cancelling and modifying projects.
Regular evaluation of the portfolio provides the basis for portfolio optimization and continuous improvement of portfolio management processes.
Portfolio Process Efficiency and Project Lead Time
An important thing to look at, when considering the portfolio management processes, is the process efficiency of the portfolio (or the process efficiency of the project life cycle). The cycle efficiency of a process is calculated by dividing the value-added time by the total lead time of the process. A low process cycle efficiency is a symptom of waste.
As we can see in Figure 8-5, the total project lead time comprises several components.
Figure 8-5: Project Life Cycle Process Efficiency
At the level of the project portfolio, we can calculate the efficiency of the project life cycle process by dividing the Project Cycle Time (the time for executing the project) by the Total Project Lead Time (in-system time).
Often, the time for non-executing life cycle phases is a large part of the total lead time. The low cycle efficiency is a symptom of waste. To reduce the lead time, we need to shorten the non-executing phases.
The project execution phase (and the other phases in the project life cycle) also has its own measure of cycle efficiency - the ratio of the value-added time to the cycle time. We address this issue in the Managing Creation and Absorption chapter.
The average lead time of the individual projects is affected by the work in progress limit (or the lack of limit) for each portfolio.
Sequencing projects does not imply that we need to wait for one project to finish before starting another. Depending on our capacity and the target average lead time, we will have a certain number of projects in progress at the same time.
Little's Law is a theorem that provides a principle for evaluating the efficiency of queuing systems. It defines the relationship between the long-term average number of customers in a stationary system (L), the long-term average effective arrival rate of the customers (λ) and the average time that a customer spends in the system (W).
In a stable system, the departure rate (throughput) is the same as the arrival rate.
The formula of the law is: L = λ W
In projects, the relationship is between average number of jobs in the system (Work in System or in-system inventory), average processing rate (throughput) and average lead time.
Work in System (WIS) is the number of jobs in the system (waiting to be serviced or being worked on).
Processing Rate is the number of jobs that are processed and depart the system for a given period.
Lead Time (time in system) is the time a job spends in the system, including queue time (waiting to be processed) and the processing time (cycle time).
The project lead time has important implications for project success due to the Cost of Time. With the help of the law, we can determine the average number of projects in the system that is needed to achieve a certain average lead time. We can define the project processing system as the time and processes framework, from project idea generation to project completion.
Let's say that the average processing rate of a value stream is 30 process improvement projects per year and we aim to ensure that the average project lead time is no more than 5 months.
Using the formula
Average Work in System (L) = Average Processing Rate (λ) x Average Lead Time (W),
we get 12.5 (30 x 5/12).
To ensure an average project lead time of up to 5 months, we must limit the average number of in-system projects to 12 and balance the arrival rate with the departure rate.
For an average lead time of 3 months, we have to set a limit to an average of 7 in-system projects. We achieve the longest average lead time of 12 months with an average of 30 projects in the system.
Project initiatives create a flow of transformations and improvements and support the organization’s flow of value.
The continuous improvement, major improvement and transformation projects, and the ongoing continuous improvements differ in the scale of impact, time horizon, frequency (number) and uncertainty. All these types of initiatives work as a system, and the organization needs an optimal dynamic balance between them to ensure its sustainable development.
Lean Project Portfolio Management maximizes the net value of the organization's initiatives, thus improving the capacity of the value-creating system and contributing to the maximization of the organization’s flow of net value.
Lean Project Portfolio Management differs from the traditional portfolio management:
- It’s managed in a decentralized way at both organizational and value stream levels.
- It uses bidirectional objective setting, and incremental and adaptive planning and funding.
- Resource allocation is fast and flexible and funds value streams that incrementally fund Minimum Viable Projects and ongoing improvements.
- Lean portfolio management uses strategic and value stream objectives and KPIs or Objectives and Key Results (OKRs) as project objectives.
- Progress measurement is value-based.
- Stable long-lived teams which are subsets of the value stream teams are preferred.
Project selection should be based on alignment, feasibility, viability, importance and priority.
Minimum Viable Projects reduce waste and increase the probability of success.
Using scoring models to evaluate project importance is problematic because this can lead to biased decisions.
An appropriate measure of project importance is the risk-adjusted expected net value.
Prioritization provides allocation of resources to individual portfolios, including allocation by strategic themes, and project sequencing.
Allocation is a bidirectional process. It should be completely decentralized at the value stream level and aligned with the capacity.
Schedule prioritization must be based on the Cost of Time.
When the postponed execution costs for the individual time periods and the durations and/or the reference points for postponement are homogeneous, we should apply the SJF, HDCF and WSJF rules to sequence projects. When these parameters are not homogeneous, to identify the correct priority, we have to compare the total costs of postponed execution of the individual sequencing combinations.
Schedule priorities depend on the time horizon and the trends in the time-related costs.
The portfolio should be reassessed and altered following regular feedback loops.
To reduce risk and waste, short feedback loops on project progress must facilitate incremental funding.
Key aspects of portfolio management are ongoing project coordination and synchronization and continuous improvement of portfolio management processes.
The portfolio process efficiency should be improved by reducing the time for non-executing project life cycle phases. The average project lead time can be shortened by limiting the Work in Progress.
We'd love to hear from you.
Share your thoughts in the comments below!
 Reinertsen, Donald G. (2009). The Principles of Product Development Flow: Second Generation Lean Product Development. Celeritas Publishing