aka: Don’t Cripple your Portfolio Agility
The Scaled Agile Framework (SAFe) provides a very powerful portfolio model that allows companies to articulate, steer, and capture the benefits of technology investments. Like the rest of SAFe, however, there is a lot of content and it is very difficult to understand which concepts and behaviors are most important to receiving benefit.
To this end, I have captured the ten essential behaviors I believe are necessary in order to allow good Lean Portfolio Management (LPM) behaviors to take root and flourish in a company. In the same way that I use the ten Critical ART Success Factors to diagnose why a single train is struggling, this is the list I look to to understand why a portfolio is struggling to capture the benefits of agility.
- Articulate a clear and unambiguous strategy
- Identify a balanced portfolio leadership team
- Tie clear outcome-framed goals to strategy
- Map value streams and provide clear visions
- Visualize and publish organization structure
- Maintain clear, well-respected guardrails
- Broadcast the kanban and roadmap
- Treat epics as experiments
- Ensure the APMO serves the portfolio
- Reject work and limit strategic WIP
Articulate a clear and unambiguous strategy
It is impossible to decentralize the decisions associated with a portfolio when the underlying strategy is unclear, unarticulated, or not distributed in a way the people making the decisions can utilize it. The first step in any effective leadership, much less portfolio leadership, is to expose the strategies along with the decisions and backgrounds that underlie those strategies. this allows individuals in various levels of leadership across the entire portfolio to begin making faster decisions that remain aligned with the desires of the senior executive team that sets the strategies. In SAFe, this takes the form of clear strategic themes, made visible across the portfolio.
Identify a balanced portfolio leadership team
As the portfolio begins to decentralize many decisions, the remaining decisions tend to become higher-stakes and carry increased risk exposure. For these decisions to be made quickly and effectively, the leadership team of the portfolio needs to be able to rapidly incorporate a wide variety of viewpoints covering as many perspectives as possible. This requires that the portfolio leadership team includes technology leadership, business leadership, HR perspectives, architecture perspectives, and finance representation, at a minimum. It also implies that each of these roles need to be a full participating member of the portfolio team, and that lower-level executives not be overturned when they raise important concerns relevant to their area. The portfolio leadership team should also be able to rapidly gain the perspective of the leaders within the portfolio to ensure that they are consistently capturing opportunities that arise from the work that is already being done. In SAFe, this informs the composition of the LPM function as well as the attendance of the portfolio sync.
Tie clear outcome-framed goals to strategy
Often, strategies become poster art and lose any of the meaning that led to them. One of the more effective tools for ensuring that the strategies retain meaning is to tie a set of quantitative and qualitative outcomes to them, allowing the leaders across the portfolio to align around improving those metrics. In SAFe, these are often captured in the form of Objectives and Key Results (OKRs), which provide both the aspirational outcome and the quantifiable targets that go with them. The most common failures when using strategic themes are to 1) have them represent work activities rather than the benefits of doing that work, 2) to have them be very narrowly specified for individual groups, so that each group has its own individual OKR and does not have a need for collaborating with other groups, or 3) have them be so long or obtuse that nobody reads them.
Map value streams and provide clear visions
Many transformations get started by picking a convenient group and calling it an Agile Release Train. In the same way, companies often start with LPM without having shaped the underlying organization into a form that is compatible with continuous funding and managing by vision and purpose. It is crucial to perform a value stream mapping exercise based on the current organization, the strategies in play, and the desires for the longer-term shape of the organization so the funded investments will flow effectively in an LPM environment. Ideally, this exercise also results in the design of the Agile Release Trains (ARTs) that will make up each of these value streams, but at a minimum the current-state organization needs to be mapped and well understood.
Visualize and publish organization structure
Once the value stream map is complete, it is important to ensure that the organization is aware of the results and the implications of future state goals. Too often, the value stream mapping is done in a small group, the results are read out to leadership, and then the highly useful visualization that was completed is never looked at again. SAFe provides very effective tools for this communication in the form of the portfolio canvas, value stream canvases, and Agile Release Train canvases. They should be completed, maintained, and published prominently in ways that leaders use them to make decisions and contributors use them to understand context. In effective portfolio implementations, they often form the foundation for the executive reporting decks which will follow a similar form while providing fresh, relevant information every two weeks.
Maintain clear, well-respected guardrails
One of the most important considerations for decentralizing decisions is providing a useful framework to support those decisions being made effectively and consistently across the portfolio. When dealing with large sums of money that are being entrusted to individual product managers, it is even more crucial that the long-term concerns of the Enterprise are reflected in how those decisions are being made, even while those product managers are pursuing their local goals for their individual products and solutions. SAFe introduces a useful set of guardrails that ensure that short-term needs are balanced against long-term goals, and that the business owners who are responsible for the financial outcomes and sustainability of the solutions are involved in the key decisions on at least a quarterly basis. Once these guardrails are established, it is important that aligning to them is perceived as a mandatory behavior, and becomes part of the culture of leading value streams and products.
Broadcast the kanban and roadmap
Visibility of both work and future-state desired outcomes is important for people to maintain alignment, and this is not possible if they are unable to see or easily access the current state of in-flight work. The collection of canvases referenced above is an important part of this, however people also need to be able to see the list of epics that are currently under consideration, active, and, most importantly, canceled. This allows individuals to know where energy is expected to be focused currently, and allows them to push back or refuse to do work on things that are not yet approved or have been explicitly disapproved. Sharing this roadmap of epics can be inspiring to people, and provides them context for the larger organization when people come with requests about work that aren’t part of the individual’s local scope.
Treat epics as experiments
The portfolio as represented in SAFe is a portfolio of solutions and the investments being made into those solutions. Unfortunately, it is too often treated as a project portfolio or a work portfolio, losing much of the value of the decentralization and limiting the impact agility can have for an organization. One of the most common errors is treating epics as an equivalent to project, or treating them as a container of work where the epic is finished when all of the features are finished. Instead, the portfolio leadership team needs to be dedicated to focusing on epics as large-scale strategic experiments that have a named hypothesis and clear pivot/persevere points identified to guide creation of a minimally-sufficient MVP. At this point, the leadership team needs to help evaluate the results of the MVP and then transition that epic out of the portfolio to become part of a ART vision or the foundation of a new value stream with its associated ARTs.
Ensure the APMO serves the portfolio
When much of the work of a portfolio is managed from a central location, that central group tends to become a significant control and coordination entity because they need to maintain alignment across a very diverse group of projects that are all in different phases. Once LPM is up and running reliably, the challenge of doing that alignment work is significantly reduced, especially given the continuous funding and clear visions associated with each group. When the organization is successfully designed to be aligned around value, the work required from the core group is significantly reduced yet again. As this transition occurs, the Agile PMO needs to ensure it is continuously shifting more and more into a service stance, with a mandate of removing friction from the rest of the organization. As described in SAFe, this behavior increasingly becomes one of “operations” rather than “management”.
Reject work and limit strategic WIP
The final critical concern of the Lean Portfolio is ensuring that it eliminates significant amounts of work. One of the most important considerations of strategy is providing focus and a way of challenging all work that does not advance the strategy. Unfortunately, many of the lean portfolio implementations I’ve seen have the eventual approval of work is a foregone conclusion, and it feels like going through the motions. In practice, a significant percentage of items in the funnel should be discarded before advancing, and the approval rate of items that hit the next two stages should definitely be well below 100%, perhaps below 50%. While it is challenging for executives to consistently refuse work, especially when it is coming from peers with important titles and high-value businesses they represent within the company, it is necessary for a company to move forward at survival speed. This part of the role of executive is “To kill the good ideas so the great ideas can thrive”.
In summary…
The hardest part of LPM is getting started. If you are successful in instilling the ten behaviors above, you will have a very strong foundation in place. Once it’s running, the system gains momentum and continues running very effectively, allowing normal continuous improvement activities to smooth out any other challenges that arise. From there, the momentum quickly helps this new way of working become the culture for how technology investment money is managed.
