Agile software development and traditional cost accounting don’t match.
—Rami Sirkia and Maarit Laanti 
Lean Budgets provide effective financial governance over investments, with far less overhead and friction, and supports a much higher throughput of development work.
When implementing Agile at scale, many organizations quickly realize that the drive for business agility through Lean-Agile development conflicts with traditional budgeting and project cost accounting methods. As a result, moving to Lean-Agile development—and realizing the potential business benefits—is compromised, or worse, blocked entirely.
Every SAFe portfolio operates within an approved budget, which is a fundamental principle of financial governance for the development and deployment of business Solutions. Figure 1 illustrates the Enterprise strategic planning process that creates each portfolio’s budget.
The SAFe approach to budgeting is significantly different than traditional methods. This approach provides effective financial control over all investments, with far less overhead and friction, and supports a much higher throughput of development work. Figure 2 illustrates the transition and highlights the three main steps in implementing Lean Budgets.
The Problem of Traditional Project Cost Accounting
Before we discuss how to implement Lean budgets, it’s critical to understand the problems caused by traditional project cost accounting.
Project Budgeting Creates Multiple Challenges
Figure 3 represents the budgeting process for most enterprises before they move to Lean budgets. In this example, the enterprise is organized into four cost centers. Each cost center must contribute budget or people (the primary cost element) to the new effort.
This traditional budgeting model creates several problems:
- Slow, complicated budgeting process – Many large technology projects are subject to siloed organizational structures, and therefore require multiple cost centers and functional managers to fund a single project.
- Lower fidelity decisions – Individuals and teams make poorer choices when they are forced to make fine-grained decisions too early in the ‘cone of uncertainty,’ when the least amount of learning has occurred—and there’s no time for validating assumptions.
- Temporary teams lower overall performance – People are assigned to a project on a temporary basis and then return to their functional silo for future assignment to a new initiative. This hinders learning, employee engagement, and overall organizational performance.
- Waiting on specialists causes delays to value delivery – Traditional project teams focus on individual skills, and it’s common for one project to block another while waiting for key personnel with specialized skills. If a project takes longer than planned—which it often does—many people will have moved on to other projects, causing further delays and lower quality.
- Full resource utilization is favored over a fast flow of value – In the pursuit of theoretical efficiency, everyone is assigned to 100% capacity, often to multiple projects. However, Reinertsen notes that “operating a product development process near full utilization is an economic disaster” . This disaster is a result of long queues, project delays, and high variability between forecasted and actual, time, and costs. A study by Adler concludes that “if managers had reduced their planned utilization rate to 80%, they could have reduced development times by 30% or more.” .
The Project Funding Model Impedes Adaptability
Once the project is underway, the challenges continue. The needs of the business change. However, because the budget and people are fixed for the initiative’s duration, the result is an organization that is unable to change the plan without the overhead of re-budgeting and re-allocating personnel (Figure 4).
Project Delays Happen. Things Get Even Worse.
Often, work will take longer than planned because of new learning, insights, and opportunities. Further, even when things go well, stakeholders may want more of a specific feature. Many organizations manage change through a Change Control Board (CCB), which again, adds more delay. The project model also hinders cultural change, transparency and solution development progress (Figure 5).
Product development also requires innovation, and we cannot innovate without takings risks . Because innovation contains a higher degree of uncertainty, it’s challenging to estimate these types of projects.
When a schedule overruns for any reason, it’s necessary to analyze the variances, re-plan, and adjust the budget after getting approval. People are scrambled to different projects, resulting in adverse impacts on other projects. Now, the ‘blame game’ starts, pitting project managers against each other, and financial management against the teams. The ultimate result is information hiding, loss of productivity, and lower employee morale.
Beyond Project Cost Accounting with SAFe
SAFe provides a different approach to budgeting which reduces the overhead and costs associated with traditional cost accounting, which empowers people through Principle #9, Decentralized decision-making. With this new way of working, portfolio level personnel no longer plan the work for others, nor do they track the cost of the work at the project level. There are three main steps, as described below.
1. Funding Value Streams, Not Projects
The SAFe portfolio budget funds and nurtures a set of development value streams, which delivers one or more business solutions. So, the first step of Lean budgeting is to give each value stream a budget as Figure 6 illustrates. A set of Lean Budget Guardrails support these budgets by defining the spending policies, guidelines, and practices for a specific portfolio. Guardrails, like any good governance, enable increasing autonomy to the value streams.
Funding value streams vs. projects delivers several benefits:
- Empowers local content authority, which moves decisions to where the information lives, enabling faster and better decision-making
- Improved clarity of spending through value stream budgets
- Better visibility to large business and technical initiatives (portfolio business and enabler epics) through the portfolio Kanban
- Knowledge workers working in long-lived value streams are simply more productive than temporary project teams
- Self-organization allows moving people to the most critical work without escalation to management
- Better management of the budget while simultaneously providing more autonomy to Agile Release Trains (ARTs) and Solution Trains
Figure 7 shows that in most cases, the people expenses across a Program Increment (PI) are fixed or easy to forecast. Moreover, features that take longer than expected do not change the budget. As a result, all stakeholders know the anticipated spend for the upcoming period, regardless of what features are implemented.
2. Guiding Investments by Horizon
Just as a personal finance portfolio balances investments in different asset classes (e.g., stocks, bonds, real estate), SAFe balances different horizons of investment. Get this balance wrong, and you can starve the future by over-investing in today, or you can miss near-term opportunities while pouring too much money into the future.
The ‘Three Horizons of Growth’ framework shown in Figure 8 and featured in the book ‘The Alchemy of Growth: Practical Insights for Building the Enduring Enterprise’  provides a structure for companies to manage investments in growth without neglecting performance in the present.
Adapted, in part, from the McKinsey horizon model , the SAFe investment horizon model in Figure 9 highlights spending allocations for solutions that are created by the individual value streams. This helps enable value stream owners and fiduciaries to make more informed decisions, and it helps align the portfolio with strategic themes while promoting overall health and growth.
Horizon 3 (Evaluating): Horizon 3 investments are those dedicated to investigating new potential solutions. Often, an Epic is created to start the initiative, and an MVP is defined to test the benefit hypothesis. Investment continues until the initiative is stopped or allowed to continue to horizon two.
Horizon 2 (Emerging): Horizon 2 reflects the investments in solutions which have emerged from horizon three. Since these are promising new solutions, the business is willing to make ongoing investments in excess of the current return. However, if the decision is made to stop the program, it’s likely that some modest investment is necessary to decommission the system, as it has likely made its way into the internal and external business ecosystem.
Horizon 1: Horizon one reflects the desired state where solutions deliver more value than the cost of the current investment. These solutions require ongoing investment to maintain and extend functionality. This investment profile often consumes the majority of a portfolio budget. For convenience in reasoning about these investments, horizon one is further subdivided into two profiles:
Investing: These investments reflect solutions that require significant ongoing investment. This may be due to market or solution immaturity, changes to the market or technology, or new windows of opportunity to be addressed.
Extracting: These investments typically represent stable solutions that are delivering great value with a relatively lower need for additional spending. Investment in these solutions is managed to assure continued value, profit, and cash flow, enabling funding of emerging solutions.
Horizon 0 (Retiring): All solutions eventually meet end-of-life. Horizon 0 reflects the investment needed to decommission a deployed solution.
The size of the bubble in the figure represents the current level of investment in a solution. This is often supported by details captured in the value stream canvas.
Value stream leaders must learn to manage all four horizons simultaneously. After all, value streams must dynamically evolve solutions, introduce and retire solutions, manage technological change, and respond to market demands. In addition, on occasion, entire new value streams must be created, and others may be retired. These choices are informed by portfolio budget guardrails, which help determine how much budget should be invested in each investment horizon.
3. Applying Participatory Budgeting
Most organizations will generate more good ideas than it can fund, resulting in a portfolio prioritization challenge. Participatory Budgeting (see Figure 10) is a technique for solving this challenge while ensuring that value streams receive the necessary funding. This enables LPM fiduciaries to collaborate with Business Owners and other relevant stakeholders to right-size the investments in value streams and helps manage the approval process of epics in the portfolio Kanban.
Since the participants can’t fund all of the items, they must work together to identify the best investments. Even more importantly, participants from different value streams must collaboratively pool their budgets to support initiatives that no single value stream can fund by itself. The discussions from this collaboration allow participants to make better choices about epics to optimize value delivery across the portfolio.
Participatory Budgeting does not require funding each value stream to the full requested amount. Indeed, one of the goals of this model is to provide leaders with the forum needed to share data that can help the portfolio optimize its budget choices.
Learn More Special thanks to Rami Sirkia and Maarit Laanti for an original white paper on this topic, which you can find here.  Reinertsen, Don. Principles of Product Development Flow: Second Generation Lean Product Development. Celeritas Publishing, 2009.  Getting the Most Out of Your Product Development Process (http://www.bcf.usc.edu/~padler/research/HBR_prod_dev_proc.pdf)  Baghai, Mehrdad, and Coley, Steve. The Alchemy of Growth: Practical Insights for Building the Enduring Enterprise. Basic Books, 2000.
Last update: 17 December 2019