It’s not too hard to get a senior executive to buy in Lean, at least on a superficial level. Reducing waste, building quality in, nurturing a culture for respecting people, encouraging innovation, focusing on the lead times and relentless improvement – yes, sounds good! The essence of flow, the changing role of leadership … hmm … those tend to require a bit more digesting.
Who wouldn’t like to be Agile? Better software faster? Yes, please! And when you go wrong, you can pivot fast and again you are back on track towards success. Moreover, all seem to practice Scrum or Kanban and some elements of DevOps, so for any 21st century SW executive it’s obvious that teams need to be agile.
When your organization grows to several SW teams, you want agility at scale, too. Scaled Agile Framework® a.k.a. SAFe® gives guidance for how to do it. Essentially, SAFe suggests that you identify Value Streams, onboard the teams into Agile Release Trains, set up periodic evaluation checkpoints with Program Increments, and arrange backlogs and governance around this structure. Scaling up agile to the Program level is a highly impactful change after which projects cease to exist and teams perform continuous development in cadence. It’s no small feat to succeed with that change but it’s still not enough – Portfolio Management needs to become Lean and Agile, too.
Agile Portfolio Management requires fundamental changes in leadership
I have seen the Lean-Agile transformation somehow paralyzing at Portfolio level in some large organizations. Portfolio level requires most change in senior leaders’ own work, so the leaders who have been driving the Lean-Agile transformation now become targets of their own change. It is the time for them to change their own beliefs, ways of working and leadership, and these changes are quite fundamental. As a SW executive you need to be ready to:
- Eliminate the ghosts of waterfall. You need to forget projects; your portfolio consists now of Epics to be implemented by Release Trains and teams performing continuous development (and delivery). Give up the false feeling of control that you used to believe you had with stage gate milestones – those no longer exist. From now on you track the flow of Epics in Portfolio Kanban.
- Manage customers’ expectations for agility. Often customers still expect waterfall-like contractual commitments several months ahead with fixed schedule and essential content. There are typically valid reasons behind this: customers may have their own product or service launches or process changes that are schedule-driven. Agility, in turn, may be interpret by customers as the permission to iterate anything until the very end of the project. If you need to give too detailed commitments too early and still accept almost any change request that a customer makes up, you will probably end up paying contract penalties due to delays.
- Decentralize control and decision making. Challenge the management meetings and internal steering groups where you made the schedule and detailed content commitments several months ahead. You will still need to make customer commitments to get deals but you need to avoid fixing the scope too early. Accept the fact that teams make now detailed commitments, not you on behalf of them, as part of Sprint and Program Increment Planning. You need to empower Product Owners and Epic Owners to facilitate decentralized decision making. Your job is to ensure that Strategic Themes and portfolio priorities are crystal clear for all to facilitate alignment to priorities.
- Re-organize to align organization with Value Streams. Bring inter-team dependencies inside the teams where possible. Discontinue the old roles that overlap with the new Lean governance.
- Move from annual planning and budgeting to quarterly (or shorter) cycles. Committing resources to detailed annual plan goals 12 months ahead would kill agility. Goal setting and funding need to support the rhythm of Program Increments (ref. Lean Budgets). A practical start is to carry out quarterly financial planning (which is typical in many enterprises anyway) and have 12 weeks’ Program Increments which roughly match with quarters. However, 12 weeks tends to be a bit too long planning period from the Program Predictability point of view. Often a better solution is to choose any preferred Program Increment duration (my favourite is 8 weeks) and have only ‘directional’ target setting as part of the quarterly business planning. That is, business and SW planning don’t need to happen exactly in sync but business plans must not force the SW teams to make waterfall-like commitments. OKR (Objectives and Key Results) goal setting framework used by many Silicon Valley companies provides one good approach for agile (business) goal setting.
- Prioritize the portfolio continuously in the cadence of Program Increments. You should have a Roadmap spanning several Program Increments ahead but that’s not a committed plan. You need to re-evaluate the portfolio priorities at Program Increment boundaries and define the business objectives for one Program Increment at a time.
- Change your own work. Align your own and your leadership team’s schedules with the rhythm of release trains and teams.
The operational changes above make Portfolio Management more Agile but how to make it Lean?
Lean Portfolio Management maximizes the value of the portfolio
The goal of Lean is to deliver the maximum value to customers in the shortest sustainable lead time. Lean-Agile Portfolio Management plays a special role in maximizing the value of the portfolio and the flow of that value that an enterprise can generate.
Interestingly, while maximizing value-add over waste is in the very heart of Lean, it seems to be difficult to articulate the value of a business case so that it stands up to scrutiny. What is ‘value’ after all, what makes an Epic ‘valuable’? An Epic creates value when it results in software or service that provides benefits that are appreciated. It depends on the role and context of individuals how valuable they perceive the benefits to be.
In my experience, most organization struggle with effective prioritization. The larger the organization, the more there are good initiatives to prioritize and more political games ongoing. Dependencies – technical and organizational – make decision making a nightmare, sometimes resulting in endless discussions without clear conclusions. Each individual business case alone might look attractive but if you choose to do ‘Epic A’ what is the opportunity cost for ‘Epic B’?
Systems Thinking helps in portfolio prioritization
Systems Thinking needs to be applied in Portfolio Management, too, to prioritize Epics effectively. First, Lean Portfolio Management team needs to build a shared understanding on the Strategic Themes, the key goals of the organization. Then the logic how the proposed Epics create value and contribute to these strategic goals is described. The emerging model will make the expected benefits and dependencies explicit and allow studying the value of Epic proposals from the systemic angle.
Value Creation Model™ (VCM) provides a practical approach to study the impacts and relationships of Epics. It is a causal loop model that can be applied to illustrate the system dynamics for how an investment creates value. It describes the value we want to create and the factors that affect value creation as well as their interactions. The picture below shows a simple, early draft example exploring the benefits and the overall value creation logic of a Customer Relationship Management (CRM) system. This text book example reflects the real-world experience well: often the benefits – value – is created ‘outside’ the actual IT system as a positive impact to business processes. A portfolio level Value Creation Model™ may have some hundreds of nodes. Each node in the model is a potential metric and a leading indicator for the key goals of the organization. Thus, Value Creation Model™ also gives us the metrics we need to measure and lead the realization of the expected benefits.
Reproduced with permission from © 2016-2017 Qentinel / Esko Hannula. All rights reserved. Original picture graphic found at “Three Skills of Advantage”, page 46.
Building a Value Creation Model™ for a Value Stream as a team work is beneficial in many ways. It makes the goals and the paths to achieve the goals explicit. A part of the modelling process is to identify the Value Paths – the causality chains that have the largest value creation potential or the highest risk of loss. Thus, the Epics that contribute most to the Value Paths should have the highest priority. Value Creation Model™ can nicely complement the proven SAFe practices by bringing Systems Thinking to Portfolio Management.
In conclusion, Lean-Agile Portfolio Management aims at maximizing the flow of value. Portfolio Management needs to live and breathe the rhythm of Agile Release Trains and teams. In order to maximize the value of portfolio flow, we need to apply Systems Thinking and understand the benefits and the overall value creation logic of the portfolio. Value Creation Model™ is a tried and true tool that suits well also for Portfolio Management purposes.
Juha-Markus Aalto leads product development of digital services at Qentinel. He is SAFe® Contributor and acts as SAFe Advisor in the large-scale Lean-Agile transformation programs supported by Qentinel.