SAFe 2: SAFe Portfolio Management

“Innovation is saying no to 1,000 things”

 – Steve Jobs

Portfolio Management is the highest level planning process within an enterprise and includes the definition, refinement, prioritization and funding of business strategy initiatives. These initiatives are referred to as epics in SAFe, and once approved at the portfolio level, are translated into implementable features that can be realized by delivery teams. Portfolio management refers to the definition, prioritization and management of business initiatives (epics) that require work at the program and team levels. This is accomplished via a sequence of interconnected Kanbans that operate continuously, and supply delivery teams with a steady flow of ready work.

Before we start, here are the basic definitions used in the SAFe organizational framework:

There are a number of specific steps in the Portfolio Management framework:

The primary output of this set of steps is a Portfolio Backlog, that is, a ranked list of business initiatives (epics), sufficiently defined that they can be translated into product features by program teams. This flow is not intended to run as a single batch for all epics. This process should run continuously, with different epics at different stages within the process.

Strategic Themes and the Product Portfolio

A product portfolio is the collection of all products or services offered by a company. Ultimately, the goal of the portfolio management level in SAFe is to maximize the financial performance of the portfolio. All major corporations perform some sort of portfolio analysis. SAFe brings agility to portfolio management via the Portfolio Kanban, which introduces lean principles of  Transparency, WIP limits, Pull and Flow. Additionally epic priority setting is done objectively using WSJF.

SAFe uses the term Strategic Themes to describe the strategy associated with a portfolio. Large, multi-business, organizations may have multiple product portfolios, and each portfolio may have its own strategic themes. There is no standard way of structuring portfolios and themes, and organizations should adopt what works best for them. If we take Microsoft as an example, the major elements of their product portfolio might be listed as follows:

Continuing with the Microsoft example, and singling out Windows as one Value Stream within the portfolio, the breakdown from Portfolio to Epics to Product Features might look something like the diagram below. Remember that Epics (like the Edge Browser for example) are things that may require multiple PI’s to deliver to the market, whereas features are things that are produced in a single PI time-box. The only hard and fast rule in SAFe is that features are things that are produced in a single PI, anything bigger than that is by definition, an epic.

Portfolio Analysis to Determine Investment Strategy

Each product or program in a portfolio serves some strategic objective (Strategic Theme in SAFe), and generates revenue for the company. Organizations conduct analysis of their portfolios to determine priorities and investment levels. Microsoft spends about 14% of its revenue in R&D (approx. $14B in 2017). The question is how to allocate those funds among the items in its broad product portfolio. No specific analysis methodology is prescribed by SAFe, though the McKinsey Investment Horizon Model is referenced on their website – this describes how investment decisions are made based on where a product is in its lifecycle. Another popular portfolio analysis technique is the BCG Growth/Share Matrix. This is a framework for allocation of resources on the basis of the attractiveness of their market and their own level of competitiveness. It is designed to bring discipline to the determination of investment levels in a portfolio of businesses, and help decide which businesses to grow, run for cash, fix or sell. The Growth/Share Matrix can be applied to a portfolio of businesses or a portfolio of products with in a business:

In a world where markets are increasingly unpredictable and subject to sudden disruption, companies need to be able to quickly shift resources among business units and products.

For Microsoft, Windows might be considered a Cash Cow, but a very large one ($20B per year) that should be defended at all costs via an appropriate level of investment and continued innovation. Azure might be considered either a Star or a Question Mark, as it is currently number 2 behind Amazon’s AWS in the in the growing cloud solutions market. Whatever the correct positioning of these products within the portfolio, deciding appropriate funding and investment levels is a crucial step to in the creation and ranking of an epic backlog.

Epic Refinement via the Portfolio Kanban

Product strategy and portfolio analysis are nothing new, and certainly not unique to SAFe.  What SAFe does offer is an agile method for defining business epics and a way to refine them to the point where they can be consumed by program teams. This is accomplished using a Portfolio Kanban to visualize the work, make status explicit, limit work in progress (WIP), and continuously improve the process. Different levels of epic refinement are reflected in a series of workflow states, each representing more detail than the previous one, until the final state contains those epics that are considered approved for investment and ready for consumption by program teams. Kanban is a very simple process with only 3 basic rules:

In SAFe, epics are business and technology initiatives that have a lifetime beyond a single program increment, or PI. Again, using our Microsoft example,  Windows, the Edge Browser, OneDrive and the Defender Security Suite could each be considered as epics. The recommended template for an epic includes:

• An epic hypothesis or value statement (elevator-pitch style vision statement) – describing what are we building, who is it for, and why do they need it – See Geoffrey Moore’s Crossing The Chasm for details and examples.
• A set of core features, or scope definition. Its also helpful spell out what’s not in scope.
• A Minimum Viable Product (MVP) definition,
• A list of any architectural or technology enablers required to support the required features.
• A lightweight business case – some basic knowledge of revenues vs, required investment levels.
• A ranking level vs. other epics in the portfolio using WSJF (Weighted Shortest Job First) – See Donald Reinertsen’s Principles of Product Development Flow for more on WSJF.

Epics that have achieved the above level of refinement are added to the portfolio backlog.

A portfolio Kanban system is an effective way to manage this process and track progress. The portfolio management team (epic owners and the enterprise architect) drives this process via regular portfolio management (or portfolio backlog refinement) meetings. Typical governance tasks include:

• Create/Review epic definitions
• Review epic business cases
• Identify architectural enablers
• Determine priorities, estimates and epic rankings (e.g. using WSJF)
• Update the Kanban board

The portfolio Kanban has the following states:

• Funnel – initial state for all new ideas/proposals pending review and analysis
• Review – Epic defined in terms of epic value statement (elevator pitch-style vision statement) and in-scope, out-of-scope items defined.
• Analysis – One -page business case, solution alternatives, enablers, relative ranking (WSJF), and Go/No-Go decision on moving to the Portfolio Backlog. Tools like the BCG Growth/Share Matrix might be used to support the analysis.
• Portfolio Backlog (Approved Epics)

All workflow states except Funnel are WIP-limited. WIP limits should be set to encourage completing one workflow step before starting new work for that state, that is, to optimize the flow of epics across the Kanban board. WSJF is continuously refined across all Kanban states as more details are discovered. Like any other agile governance event, the Kanban review process operates on a regular cadence to ensure a steady flow of progress.

Portfolio Management Checklist and Summary

• Portfolio Kanban system setup to facilitate the analysis and refinement of portfolio epics. PPM team uses lean governance principles – transparency, WIP limits, and flow.
• PPM (Program Portfolio Management) team in place to steer the transformation of business strategy into tangible artifacts  (epics) that can be consumed by program teams for implementation.
• Cadence established for portfolio management processes.
• Epics are defined in terms of value statements and scope.
• Epics are ranked objectively using WSJF or equivalent technique.

Once we have a properly defined and prioritized portfolio backlog, program teams can engage to elaborate epics into product features, and from there feed the PI planning process.

SAFe uses the term Value Stream to describe all of the people, processes and flow of information from concept to cash for any individual item in the portfolio. In our Microsoft example, we might refer to the Windows Value Stream, or the Azure Value Stream.

Portfolio Review

We have mainly been discussing Portfolio Planning – how the epic intake process works from the proposal/idea funnel to the Portfolio Backlog. Once implementation has begun on an epic, ongoing review is required in order to decide whether to continue development, modify or potentially terminate further development. During Portfolio Planning we conducted a basic business analysis and also ranked epics using WSJF. As epics are developed and deployed,  factors affecting WSJF rankings will change, and at some point WSJF should provide guidance on whether continued development makes business sense.  Therefore a regular (cadenced) portfolio review is needed to monitor updated epic WSJF rankings, and other appropriate data such as OKRs, and to make decisions on whether to continue or pivot on epic development.

References

• Competitive Strategy: Techniques for Analyzing Industries and Competitors, Michael E. Porter, The Free Press.
• Crossing The Chasm. Geoffrey Moore, Harper-Collins.