Product and project managers often face questions like "What is the status?", "When will it be done?", and "How many epics or features can we deliver by then?". These are crucial queries, yet difficult to answer, especially for large, complex projects that rarely stay on schedule.
Traditional planning methods often lead to frustration and missed deadlines, as teams are pushed to adhere to rigid plans that frequently fail. As the pressure mounts, work piles up, and chaos ensues. This is where Portfolio Kanban can offer a fresh approach, providing a more flexible and accurate way to plan and forecast at the portfolio level.
We’ll explore how Portfolio Kanban can be used for forecasting, compare traditional and Kanban approaches to portfolio planning and tracking, and explain how Monte Carlo simulations can enhance forecasting accuracy.
Traditional Project Management and the Challenge of Estimation
In traditional project management and many Agile frameworks, portfolio planning and tracking often rely on a set of tools such as burn-down charts, project bubble charts, and Gantt charts. These tools are based on the assumption that we can accurately estimate the time required for future work items. However, when it comes to knowledge work, such estimates are often inaccurate, and the more complex the project, the less reliable the estimates become.
The Problem with Estimation
This reliance on estimation leads to a frustrating cycle. Despite our best efforts, estimates are rarely accurate, which inevitably results in delays and increased pressure on teams. So why do we continue to rely on them? The answer lies in the need to determine the project’s scope and timeline to estimate its budget. This process often involves weeks of detailed estimation, providing a false sense of clarity about when work will be completed.
The Consequences of Inaccurate Estimation
Once the budget and workload are "known," the project is added to the portfolio monitoring system. If work items take less time than estimated, everything is fine. But if they take longer, which is often the case, the project status turns yellow or red. This is when the pressure mounts and teams find themselves working weekends to meet unrealistic deadlines.
Kanban as a Solution for Portfolio Forecasting
A Lean Approach to Planning
Kanban offers a more effective approach to planning and prioritizing future work. As a lean planning tool, Kanban helps organizations improve project management by visualizing workflows, breaking down projects into manageable tasks, and using historical data to guide decisions.
Flexible Commitment and Data-Driven Decisions
Instead of committing to a fixed scope, time, and budget, Kanban encourages a more flexible approach: commit to a shorter timeframe, complete as much work as possible, and use the actual results to forecast the remaining time and cost needed for the project.
Practical Application of Kanban in Portfolio Forecasting
Implementing this approach is not without challenges, as it requires buy-in from both vendors and customers. However, it is a more realistic way to manage projects with many unknowns, offering protection for both sides.
Imagine a large project with an initial estimate of 25 months. Instead of committing to the full duration, you propose to the customer that you work for five months, consuming one-fifth of the budget. After this period, you assess:
- Progress: Have you completed approximately one-fifth of the work? (It’s common to be slower at the beginning.)
- Budget: Have you used one-fifth of the budget?
- Risks: Have any new risks emerged?
- Relevance: Is the project still relevant for the customer, or have market conditions changed?
If your initial estimates were accurate, you’d be on track. But since this is rarely the case, both you and the customer can now reassess the project. You might need to inform the customer that the initial estimate was off and that completing the project would require double the budget. The customer can then decide to proceed, adjust the scope, or halt the project, preserving most of the budget.
10 Years Kanban Experience In 1 Free Book.
Project Manager's Guide to Kanban
This approach protects both parties, as it limits potential losses and allows for adjustments along the way. However, accurately determining the project's time requirement is crucial, and this is where project and portfolio management tools come into play to support effective Kanban-based forecasting.
How to Use Monte Carlo for Portfolio Forecasting?
One significant advantage of implementing Portfolio Kanban software is the ability to leverage statistical analysis, such as Monte Carlo simulations, to forecast how much work your team can accomplish within a specific timeframe.
This analysis draws from your historical data, and the good news is that you don’t need an extensive amount to get started. While having more data will enhance the accuracy of your forecasts, even a set of 20-30 completed items is sufficient to begin making informed predictions.
It's important to note, however, that these forecasts are based on probabilities rather than certainties. We must acknowledge that project management exists in a probabilistic rather than a deterministic world. This shift in perspective means that when we provide timelines, we should attach a probability to those dates.
For instance, instead of stating that a project will be completed by November 20th, we should say, "There is an 85% probability that we will complete this project by November 20th." This approach offers a more realistic view of project timelines and sets clearer expectations for stakeholders. It’s a different statement.
Businessmap is the most flexible software platform
for outcome-driven enterprise agility.
In Summary
Forecasting on the portfolio level is possible when:
- You negotiate shorter projects or iterations.
- You implement a flow-based approach such as Kanban at the team and portfolio levels.
- You have some historical data to base your forecasting on.
