Scrum of Scrums and Adaptive portfolios using Now, Next, Later
An organisation’s portfolio of work represents the implementation of its strategy. While some organisations believe that strategy is “set and forget” for a 1-5 year period, smarter organisations recognise that their strategy is constantly changing, and therefore so is their portfolio.
Perception of a Roadmap
A lot of organisations ask for a Portfolio Roadmap or (“Product Roadmap”) expecting something that resembles this picture - a sequence of deliverables that can be mapped out months, or years, into the future. They like the simplicity of the sequence - they know what they’re getting. They like the order of the sequence - they see a timeline in which they can assign milestones or deadlines. But this is a false representation of how portfolios of work are delivered, and will only lead to frustration if you represent it in this way.
Truth of a Roadmap
Making work visible is the best tool that we have for managing stakeholder expectations, and representing work and priority correctly is the best way to have meaningful conversations with your stakeholders. It allows trade-off discussions, disagreement, expectation-setting, and ultimately, alignment. The diagram on the right is a more accurate representation of how portfolios are delivered. The narrow, or “pointy end” represents what we know now, and the concrete work we are doing now. We have the greatest clarity and certainty of this work. The broad end represents all of the potentiality that we might consider in the future as the priority of work. We can’t do everything in that broad end, so as it gets nearer, decisions must be made and priorities must be set.
Reality of a Roadmap
Our priority of work is informed by the decisions we make along the way, and each decision point comes from learnings that can only happen in a particular order, and at a particular time. The picture on the right is a more accurate reflection of the Roadmap of work. We constantly navigate our way through the decision tree, and the result of the decision becomes the work that is in focus at the current time. But until we have navigated those decisions, all options are potentiality. Beyond a certain horizon, there is no point mapping the potentiality, because the options become too numerous to identify and manage.
How then, can we represent an adaptive portfolio of work in our Kanban board to hold meaningful discussions about delivery priorities, and to plan and orchestrate our work in a meaningful way?
How did we use Now, Next, Later at Shell Energy?
At Shell Energy, in our transition from Projects to Products, we established a planning board that resembled a Now, Next, Later roadmap that allowed plans to be able to change. The nature of work in the organisation was a portfolio of projects that the business wanted to deliver. By moving to this style of planning board, we shifted the conversation away from projects, deadlines, milestone commitments, and scope. We broke projects into smaller, more valuable items we called Features, which could then be prioritised against each other, and delivered either in parallel or staggered as far apart as desired. Our operating model gave the business the greatest flexibility to prioritise work according to their needs. Our portfolio and our 12 month plan was in constant flux as the business navigated market changes and regulation changes.
The below image is a capture of our portfolio kanban board that we would use to discuss plans among Delivery Leads. We would use the board during Scrum of Scrums to discuss the inflight work in the current horizon to resolve immediate issues. We would then discuss the next horizon to ensure that each team was aware of their upcoming work, and that they had begun the discovery process. The third horizon would bring awareness for upcoming work that is in the early discovery stage, which usually involved architecture considering alternative approaches and having early discussions with teams. Not all projects required such lead time, but it was better to have the discussion early.
The idea of each horizon is to show a commitment level to the plan in that timeframe, based on a certainty level which we assigned probability measures - 95% firm for Now, 80% firm for Next, and 60% firm for Later. The business couldn’t be certain that the plan in the 6-12 month time frame wouldn’t be interrupted by something more urgent, and then delayed as a result.
These probability measures give us a gauge for how heavily we should invest effort in the work in each time horizon. In the Later horizon, work is almost a 50/50 as to whether it is confirmed, so little effort should be invested at this time. As the work in that horizon draws nearer, it’s important that we begin to invest time to understand it better.
Work in the Later horizon is large and lumpy and still not very clear. It resembles a project - a body of work with an apparent scope and purpose, but little clarity on which parts bring the most value. This work requires a high-level solution and breaking down the body of work into smaller, more valuable pieces, like Features.
Work in the Next horizon is a set of Features with some clarity of definition and purpose. The high-level solution is refined and understood. The priority of the Features is relatively clear.
Work in the Now horizon is the work currently being built and delivered. There is very little interruption that should delay this work, as it is very well understood, and the highest ranked value in the portfolio at this time.
You can see the range of months for each time horizon, with 0-3 months representing from now until 3 months from now, 3-6 months, and then 6 - 12 months. Anything beyond 12 months is considered too likely to change in that time frame.
We would update this board every 2 weeks during our Scrum of Scrums, where Delivery Leads would attend and discuss current and upcoming work with other teams, with architecture, with the environment manager, and so on. This rapid planning and orchestration replaced the need to have quarterly planning sessions, and remained much more current as a result.
At the start of every month, we would move each horizon forward by one month. This brought new projects into the Later horizon, and moved work from Later > Next, and from Next > Now. Having a timeframe window allowed teams to take ownership of the portfolio of work, and be able to time discovery activities to their needs.
This approach to delivering the portfolio avoided committing teams to large projects and hard deadlines. By breaking down projects into Features, it gave the business the greatest ability to finely prioritise the most valuable work.
I improve the performance of delivery teams.
