Strategy to Stories: Aligning, Planning, Forecasting, and Tracking Continuously
One of the bigger challenges I have experienced in helping organizations mature their Lean and Agile practices is going from team level agility to portfolio and organizational level agility. This is exemplified in the planning processes. Big up front planning approaches are already slow and painful in many companies. Those pain points tend to be further magnified when you attempt to move to a more continuous planning approach because of high transaction costs involved in the process. This is an area where many companies tend to hit a wall in their journey.
The problem is no different than when teams first move from long development cycles to delivering completed software every two weeks. They need to identify where they have high transaction costs in their development process and identify ways of reducing them incrementally and iteratively. A simple example of a high transaction cost in development that most people can relate to is manual regression testing. It’s very difficult to deliver software every two weeks if your regression testing is manual so it quickly becomes a pain point. Some typical pain points you tend to see at the portfolio and organizational level of planning are:
- The process is tedious and can consume significant numbers of people
- Requires that all deliverables are defined and detailed enough for the teams to estimate them
- Evaluating the impact of changes is very difficult and disruptive so plans tend not to be updated as changes occur
- Teams don’t realize they are behind until it is very obvious
- We don’t compare our plans to actuals to see how we can improve because information is scattered making it difficult to do an apples to apples comparison so the process rarely improves
- Planning artifacts are static and provide value to a limited audience rather than all involved
- No single source of truth. Information lives in multiple places making it cumbersome to update
- Limited visibility and transparency as to how plans were created and validated
- No big picture overview
- Plans are deterministic rather than probabilistic so people take them as guarantees
- Difficult to account for dependencies especially between other programs and portfolios
- Doesn’t scale
- Plans are not realistic or reliable due to all the above
To help address this challenge I created a tool that allows companies to begin addressing these pain points. Now, I am not for a moment suggesting that a tool is going to resolve all of the aforementioned challenges. I fully acknowledge there is a lot more involved. This is a tool that can assist groups and companies that have decided they want to move to a more continuous way of planning, but are hindered by much of the tedium involved and may not have the budget or sponsorship to acquire a commercial tool. It also provides a way for you to experiment with how such a process might work in a simple, inexpensive, and non-disruptive way.
The image above is an example portfolio visualizing the planned work, the date range in which the work will be completed, the likelihood of completion in that date range, and how the work aligns with higher level objectives. On the surface it is a road-mapping tool, but under the hood it does much more than create a nice picture. Some of the more prominent features and benefits are:
- Generate probabilistic data driven roadmaps that can provide for improved decision making
- Roadmap items can be generated based on fixed dates and duration or forecasted durations based on number of issues and historical cycle times so you can account for any type of work
- Drastically reduce the amount of time teams spend assembling or changing estimates via automated Monte Carlo simulation forecasting rather than using very subjective effort driven approaches
- Roadmap items can be linked to Jira epics allowing the roadmap item’s start date, durations, and completion date to automatically be updated in near real-time based on the progress of the epic’s child issues
- Provide a comprehensive portfolio or organization view via master roadmaps that aggregate any number of child team road maps
- Forecast anywhere from epic(multiple stories/issues)level all the way up to strategic initiatives
- Allow you to see how strategic and business goals align with actual execution at any level
- Evaluate multiple planning scenarios quickly, cheaply, and accurately
- Account for dependencies between roadmap items either within the team or with external teams in other roadmaps
- Provide visual indicators when completion dates have slipped beyond key threshold dates
- Highly configurable so the presentation can be tailored to each specific audience all coming from the same source of truth
- Quickly and easily determine the schedule impact of changes to existing or new priorities, team changes, and/or new dependencies
- It’s free!!!
Setting up individual team roadmaps is simple and 100% data driven. Roadmap items are entered in a single tab in simple row/column format. The image above illustrates the three different ways roadmap items can be entered and thus generated. The first item has a starting date and explicit durations. The second item is dependent on the first for its starting date, but has story counts that will be used to forecast the durations via Monte Carlo simulation. The third item is dependent on the second item and as well as an item from another team for its starting date. It has the Update Via Jira option set so the durations will be forecast based on the number of issues contained in the epic associated with the roadmap ID via Monte Carlo simulation. This makes creating and updating plans very simple. In addition the roadmap is now a living document that can update automatically in near-real time based on changes made in Jira to the related issues eliminating much of the overhead associated with keeping plans up to date.
Roadmaps can be combined into master roadmaps allowing you to aggregate plans so you can have a macro level view of your portfolios or entire organization.
In the image above there are 6 team roadmaps and a single master roadmap. The team roadmaps can be combined into a single master roadmap simply by adding the spreadsheet id of each of the team roadmaps to the Imported Roadmaps configuration settings of the master roadmap along with their name and color settings. The image below illustrates the 6 team roadmaps that will be included in the master roadmap.
The image below shows the resulting master roadmap. I have also turned on the category grouping so the alignment of roadmap items to their upper level features, themes, and objectives can be seen. Like the team level roadmaps the master roadmap updates in near real-time. Any updates to the child team roadmaps will be reflected in the master roadmap providing you an easy way to understand the progress of the work being done and how that aligns all the way up to the strategic level. All the aforementioned capabilities will help to significantly reduce the overhead involved in planning and tracking and help provide a pathway to a more continuous approach.
The LAD Roadmapping Tool is freely available to anyone who wishes to use it. Just make a copy to your local Google Drive. The complete documentation is available in the Lad Roadmapping Tool Guide. I hope you find it useful. Questions and comments are both welcomed and encouraged. Enjoy!