guidance-iconApplying Participatory Budgeting in SAFe

Without involvement, there is no commitment. Mark it down, asterisk it, circle it, underline it. No involvement, no commitment. [1]

—Stephen Covey


SAFe applies a Lean-Agile approach to financial governance that increases throughput and productivity by reducing the overhead and friction associated with traditional project cost accounting. It improves the flow of innovative products and services by fostering and empowering decentralized decision-making.

Lean Budgets require a SAFe Portfolio to establish funding for each Value Stream along with a set of Guardrails that promote the prudent application of the funding. After that, value stream stakeholders are empowered to apply that budget to advance the Solutions and implement the Epics that have been approved through the Portfolio Kanban system. As changes naturally occur, Agile Release Trains (ARTs) and Agile Teams make the decisions necessary to optimize economic outcomes measured and reported via KPIs.

Participatory Budgeting (PB) is the process that Lean Portfolio Management (LPM) uses to allocate the total portfolio budget to the value streams. PB engages a diverse group of Business Owners, technical leaders, and other stakeholders in the decision making necessary to establish and adjust value stream budgets over time. PB provides several benefits:

  • Provides leaders with insights and perspectives from multiple stakeholders
  • Creates alignment on difficult funding choices
  • Improves engagement and morale
  • Reduces implementation time and overhead

PB has four phases, as shown in Figure 1.

Figure 1. Overview of the participatory budgeting process

This article describes how PB is applied in SAFe via an example derived from a hypothetical SAFe portfolio that supports the loan origination and processing (Figure 2). The portfolio has two value streams, Loan Application and Core Banking, that manage four solutions, each developed and maintained by a dedicated ART (Figure 2).

Figure 2. Loan application and processing portfolio

Phase 1: Preparing the Content

Business Context

The first step in preparing the content is establishing and sharing the business context. This context includes the current state of the business and the current Strategic Themes and Portfolio Vision, as captured in the portfolio canvas. This information is critical, both in preparing the investment opportunities to bring to the PB event and influencing the decisions made during the event.

Additional information such as solution and portfolio Roadmaps can be useful to highlight market rhythms and market milestones. In our example, the business is facing an economic downturn, which will be reflected as a reduction in the total portfolio budget allocated for this period.

Total Portfolio Budget and Guardrails

The total portfolio budget is, of course, an input to the PB process. The portfolio budget describes the allocation of this portfolio by the enterprise, and it’s the money available to fund the value streams within that portfolio. The Lean budgets and guardrails guide how the value streams spend money. These guardrails will also help to influence the mix of investment opportunities that are brought into the PB event.

The portfolio in our example is associated with a business facing an economic downturn. Accordingly, the enterprise has reduced the total portfolio budget from 21,000K to 20,000K.  This shortfall will need to be addressed by reducing the funds allocated to the individual value streams.

Establish Baseline Solution Investment

To budget responsibly, stakeholders must understand the current investment context. Each value stream budget has two parts, as illustrated in Figure 3.

Figure 3. Two components of a value stream budget

Baseline Solution Investments (BSIs) are those costs incurred by each value stream as it develops, supports, and operates the solutions that deliver current business capabilities [2]. BSIs are used in PB to ensure that existing solutions are either funded to continue to support the needs of the business or identified for decommissioning.

BSI includes the value stream cost of maintaining and improving functionality. These reflect those costs in ongoing development and maintenance of the solution. In some contexts, these costs may include operations, support, sales, and marketing directly applicable to a specific solution, or allocated to a solution as a portion of the total portfolio costs. Awareness of these costs also helps the LPM team better assess investments designed to create future cost savings.

Typical BSI costs include:

  • Employee costs (salary and burden)
  • Contractor costs (such as outsourced development or specialized services)
  • Supplier costs (such as solutions used or integrated during development)
  • Hardware or other physical materials, licensing, and additional resource costs
  • Corporate allocations for shared services
  • Other line items required by the portfolio

The LPM team can compute the costs using a simple spreadsheet (Figure 4). Note that the ‘% Allocation’ column reflects the degree to which an Agile team within the value stream is allocated to the solution. Determining per-solution costs is done for all solutions associated with the Loan Application value stream so that the LPM team can analyze costs on a per solution basis.

Figure 4. Annual and per PI costs for the credit scoring solution

When an ART is creating a single solution, establishing baseline solution costs is straightforward; it’s merely the cost of the ART.  When an ART develops more than one, the LPM team should allocate a portion of the cost to each solution.

A portfolio may choose to decompose BSI costs into specific line items, so everyone understands the nature of the investments. It also helps to identify any solution that is taking a disproportionate amount of their budget in one area.

BSIs are allocated for a specific budgeting period. For example, if the budgeting period is six months, and the PI cadence is every three months, then a typical budgeting horizon for BSIs is every two PIs. This per-PI allocation also helps the LPM team make finer-grained adjustments to the BSIs and LPM in making investments in epics, as described later in this article.

For our example, the BSIs based on the last two PIs are summarized below and help inform the requested budget for the next two-PI budgeting period (Figure 5).

Figure 5. Baseline solution investments for the loan application and processing portfolio

Proposed Solution Initiatives

Proposed Solution Initiatives (PSIs) describe the significant development initiatives intended to improve current solutions or introduce new solutions. These epics may originate from common portfolio concerns or opportunities, or they may be specific to an individual value stream.

These include both business epics associated with end-user functionality and enabler epics associated with either developing an architectural runway that enables future business capabilities or addressing compliance, security, and technology infrastructure needs. Additional details of each epic that assist the LPM team in reviewing the epic are contained within the epic Lean Business Case and will also be brought into the PB event.

For those epics that have originated from the value stream, only those that exceed an established threshold require LPM attention (see this article for advice on forecasting the cost of an epic). This guardrail helps to balance the goals of decentralizing decision-making with maintaining reasonable financial oversight. For the Loan Application Value Stream, the epic threshold is $500K. This threshold means that any development initiatives (such as new Features or technology improvements) that cost less than $500K are generally not considered a portfolio concern and are part of the BSI.

LPM reviews and agrees upon these significant investments to promote transparency and optimize the entire portfolio (see the callout “Why do we need epics in SAFe”). An overview of the PSIs for the example portfolio appears in Figure 6.

Figure 6. Proposed solution initiatives for the loan application and processing portfolio

Business agility requires that the portfolio make the best possible investment decisions for both new and in-inflight work. Accordingly, all in-flight epics are included to ensure that continued investment is still justified given current market conditions. If new and better opportunities arise, work on in-flight epics can be stopped. Previous investments from a prior budgeting period are sunk costs and are not considered. The inflight epics for our example include (Figure 7):

Figure 7. In-flight epics for the loan application and processing portfolio


Why Does SAFe Need Epics?

The purpose of epics can sometimes feel like a paradox: Why do we need epics if SAFe promotes decentralization of decisions by funding value streams?

The answer is based on how SAFe promotes healthy collaboration and trust-based relationships. Consider a married couple. Both partners are runners. The couple also has an old car. The car is so old that it is breaking down quite frequently, and the mechanic has suggested that future repairs are not worthwhile. Accordingly, the couple has agreed that they will not make future repairs and purchase a new car in the next few months.

Let’s assume one partner purchases a new pair of running shoes (sneakers or trainers). It is unlikely that the other partner would care about the purchase because it small and part of the everyday budgeting and expense process.

However, if one partner purchased a new car without consulting the other, their relationship would likely suffer. It doesn’t matter that the couple had previously agreed that they would eventually invest in a new car. The point is that purchasing a new vehicle is a considerable, infrequent expense that should be discussed.

That’s a key purpose of a SAFe epic: it promotes healthy relationships by creating a means to discuss significant investments openly.

Establish Budget Time Horizons

Any budgeting process must establish the time horizon of the budgeting period. Since an epic can be started, stopped, or finished at any time, the timeline of an epic rarely aligns with a specific budgeting period (Figure 8).  Accordingly, each PB event should address only that portion of the epic’s cost that is likely to be consumed in the upcoming budgeting period.

Figure 7. The budget for a period includes all the work on epics during the period
Figure 8. The budget for a period includes all the work on epics during the period

For an epic that has not yet started, it generally doesn’t make sense to budget for the entire estimated implementation cost until the MVP hypothesis has been proven. It may or may not be proven, and even if it is, some of the planned investment may well be outside of the current budget period. As development proceeds, the estimated implementation cost is refined based on MVP data. Accordingly, LPM budgets the forecasted investment of the epic expected to be incurred in the upcoming budget period.

For example, In Figure 8, LPM would budget for the full cost of Epic 1 since it is expected to be completed within the two-PI budgeting time horizon. Still, only that portion of Epic 2 costs expected to be incurred in the next two PIs.

For in-flight epics in which LPM believes the continued investment is warranted, LPM also budgets for that portion of the epics cost forecasted to be incurred in the budget period. In Figure 8, these are the remaining costs associated with the Enabler epic. During PB, participants will review the total forecasted costs and restrict their funding to only the costs incurred in the upcoming budgeting period. Both of these costs are shown in Figure 9 below.

Determine Proportional Costs from each Value Stream

Epics that are split across values streams, such as the ‘Move to Flutter’ epic, are funded proportionally by the responsible value streams. In our example, this epic is forecasted to cost 800K. If the core banking solution only requires 200K, then the remaining 600K must be budgeted by the other value stream, which does the rest of that work. To support transparency during the PB, and to enable value stream budgets to be calculated following the event, the value stream cost allocations for the PSIs should be made clear, also shown in Figure 9.

Figure 9. Epic cost inputs for the Loan Application and Processing portfolio

Phase 2: Assemble the Participants

PB is most effective when a diverse and responsible set of representatives are involved. Each role brings a unique and valuable perspective to the decision-making process. Attendees typically include:

  • Product and Solution Management
  • Epic Owners
  • Enterprise and Solution Architects
  • Business Owners
  • Other relevant stakeholders, such as finance, marketing, or sales

The total number of participants is typically a function of the size of the portfolio, with more extensive portfolios requiring larger numbers of participants. Because collaborating in complex negotiations tends to break down with more than 8 participants, the participants are organized into groups of 5 to 8 people per group. To promote understanding, each group should have a mix of roles from the different value streams. Our example includes 40 people, organized into five groups of eight people.

Phase 3: Conduct the Forums

The primary purpose of PB is to provide a process to collect feedback that allows the LPM team to make a more informed decision on how to allocate the portfolio budget. For example, the LPM team could reduce investments in some baseline solutions to ensure funding for the requested epics. Alternatively, the LPM team could stop in-flight epics or defer new epics.

Before the PB forums are initiated, participants are briefed about the business context, solutions, and epics. A sample is shown in Figure 10.

Figure 10. Participatory budgeting event agenda

At the start of the PB forums, participants are given the list of BSIs and PSIs and their forecasted costs (Figure 11). In the exercise, each group is assigned the full budget to allocate. Each participant in each group is given an equal portion of the total portfolio budget.

In our example, the total budget requested for all baseline solutions and epics is 24,900K, which exceeds the 21,000K portfolio budget.  Assuming the total portfolio budget is fixed entering the budgeting cycle, exceeding i is not an option.

Figure 11. Total forecast for baseline solution costs and proposed solution initiatives

Since no one participant can likely fund any one item alone, and since the participants cannot probably support all the items, they must work collaboratively to identify the best investments. Generally, the first order of business is to fund the basis of the business context. In other words, the current baseline investment may not meet the needs of the evolving strategy, and these must be analyzed and potentially adjusted as well.

To fund initiatives, individuals in each group must pool their budgets. Because partially funded solutions and epics are candidates for termination or cutbacks, the group will negotiate to determine where to make the best investments. The discussions from this collaboration allow participants to make choices that optimize value delivery across the portfolio, rather than focusing on their value streams.

After funding or adjusting BSI, teams are encouraged to support their highest priority epics fully. This guidance avoids spreading too little money across too many initiatives that are not funded sufficiently. Full investment in a selected set of initiatives is the key to a strategy that builds on an enterprise’s strengths and creates viable and differentiated product offerings.

During a forum, a group may:

  • Invest more than the estimate if they have reason to believe the costs are too low or if they think it should be delivered sooner or with more value
  • Invest less if they have reason to believe the costs are too high or if they think the investment should be deferred to a future budget period

Figure 12 is an example of how one group allocated its budget. Highlights:

  • The group agreed to substantially reduce the introduction of new features in the channel management and loan origination solutions and thereby reducing the BSIs
  • The group preserved the full investment in the revenue-generating Facebook login and college student loan epics
  • The group deferred many of the enabler epics and maintained a less aggressive investment in the core banking microservice epic
Figure 12. One group’s budget recommendation

Phase 4: Analyze Results

The aggregated results of the PB forums are analyzed to identify suggested funding patterns. In Figure 13, for example, (derived from a real PB event), it is easy to see alignment on funding priorities:

Figure 13. An example of an output of participatory budgeting event

While the results of the PB event do not determine the budget allocations for the value streams, they provide critical data and input that the LPM team needs to make an informed decision that maximizes the outcomes across the portfolio.

Some common patterns of investment are often observed:

  • Investing a consistent amount in one or more solutions that are generating stable returns
  • Reducing investments in one solution to invest more in a different solution with a better-projected return
  • Investing in the decommissioning of obsolete solutions to free resources for innovation
  • Investing across all solutions to realize a standard set of business requirements, such as compliance or serving a new or adjacent market
  • Reducing investments across all solutions during an economic downturn

These patterns are not exhaustive, and the needs of a specific portfolio will determine how they manage solutions investments.

Given this input, the LPM team makes final determinations to implement these changes, including:

  • Finalizing the portfolio budget
  • Allocating investments to value stream budgets
  • Verifying that the investments are aligned with the Lean budget guardrails
  • Engaging with HR, finance, legal, operations, and any other affected groups

Value streams typically prepare for these changes during the current PI and implement them at the start of the next PI.

Adjusting Value Streams Dynamically within Approved Budgets

Once budgets are allocated to value streams, the value streams adjust their work dynamically to address current facts, local context, and emerging business results. Some examples might include:

  • If the epic hypothesis is disproven, the value stream will stop work on the epic. If a pivot decision results in a new hypothesis, the epic owner may create a new epic and place it in the funnel of the portfolio Kanban. In the meantime, ARTs and teams are empowered to pull the next approved epic from the portfolio backlog or the next set of approved work items from the program backlog.
  • If the epic hypothesis is confirmed, the value stream will continue working on the epic, using the budget that has been allocated to the value stream.

This dynamic structure is the essence of the Lean budgeting approach and provides a far more agile way to address budget changes and have the teams work on what matters at that time.

A final consideration about budgeting: a future or forecasted budget in SAFe is never immutable. Being Agile means that the portfolio can adjust budgets at any time necessary.


Participatory budgeting is a dynamic, collaborative process that enables LPM to gather the data and build the consensus required to invest in the best possible solutions. It is a key element of SAFe Lean Portfolio Management and is used to establish Lean value stream budgets. Lean budgets and guardrails support decentralized decision-making and empower value stream stakeholders to make the ongoing investments needed to address emerging opportunities.

Learn More


[2] In some enterprises, the total of all baseline solution costs is sometimes referred to as ‘Run the Business’ or ‘Business as Usual’ costs. In SAFe, these costs are allocated on a per solution basis to bring additional visibility into the budgeting process.

Last updated: 23 November 2020