Why Traditional Estimation Falls Short
If you've spent any time in Agile, you've likely sat through a story-point estimation session that dragged on far longer than expected. These sessions are meant to help teams gauge effort, but they often lead to debates over whether a task is a three-point story or a five-pointer. And at the end of it all, does the number really mean much?
The reality is that estimation is a guessing game. Teams try to predict complexity, effort, and risk, but software development is inherently unpredictable. The good news? There's a better way to forecast work—one that doesn’t rely on story points at all.
Enter Monte Carlo analysis. This method utilizes real performance data, rather than subjective judgments, to predict the completion of future tasks. Instead of debating complexity, teams look at real throughput data—how much work they complete in a given time—and use that to forecast delivery timelines with confidence.
What is Monte Carlo Forecasting?
Monte Carlo forecasting isn’t new—it’s been used in fields like finance, physics, and risk analysis for decades. In Agile, it works by running thousands of simulations using historical throughput data (like the number of work items completed per sprint) to predict when work will be finished.
Unlike traditional estimation, which often feels like staring into a crystal ball, Monte Carlo analysis gives teams probability-based insights. It answers questions like:
What are the chances we’ll complete this backlog in the next 10 sprints?
If we need to ship by a certain date, what’s our probability of success?
How many work items can we realistically finish in the next quarter?
Instead of a single number that suggests certainty (which is rarely accurate), Monte Carlo forecasting gives a range of possibilities with confidence levels—a far more realistic approach.
Why Monte Carlo Analysis Works Better Than Story Points
The biggest problem with story points is that they’re subjective. What one team considers a five-point story, another team might see as an eight. Over time, teams can get better at sizing stories, but the process still relies on human intuition rather than data.
Monte Carlo analysis removes this subjectivity by focusing purely on historical performance. Here’s how the transition works:
Collect Throughput Data – Instead of estimating effort, track how many user stories or tasks the team completes per sprint.
Run Monte Carlo Simulations – Feed this data into a Monte Carlo tool, which will run thousands of simulations based on past performance.
Analyze Probability-Based Forecasts – The tool generates forecasts with confidence levels (e.g., “There’s an 85% chance we’ll finish within 8-10 sprints”).
Plan With Data, Not Guesses – Use the probability-based insights to make informed planning decisions.
By using data instead of estimates, teams remove the overhead of estimation sessions and make Agile planning more realistic and reliable.
The Benefits and Drawbacks of Monte Carlo Forecasting
Like any approach, Monte Carlo forecasting has its pros and cons. While it offers significant advantages, there are some considerations to keep in mind.
Why Teams Love It:
✅ Removes Estimation Bias—No more subjective discussions about complexity.
✅ Based on Real Data—Uses past performance rather than educated guesses.
✅ Faster Planning—No need for lengthy estimation meetings.
✅ Handles Uncertainty Better—Gives confidence ranges rather than fixed numbers.
✅ More Accurate Over Time—The more data you have, the better the forecasts.
Potential Challenges:
⚠️ Needs Historical Data—If your team is new or lacks past sprint data, forecasts may be less reliable at first.
⚠️ Doesn’t Capture Individual Story Complexity. While it predicts timelines well, it doesn’t consider the difficulty of each individual task.
⚠️ Requires Understanding of Probability—Some stakeholders may struggle to interpret probability-based results.
⚠️ Works Best with Consistent Teams—If team size or workflow fluctuates often, forecasts may vary.
Despite these challenges, many teams find that Monte Carlo forecasting makes planning easier and more predictable compared to traditional estimation.
How to Get Started with Monte Carlo Forecasting
If your team is ready to ditch story points and embrace probability-based forecasting, here’s how to start:
Step 1: Gather Throughput Data
Monte Carlo forecasting relies on historical data. Start tracking:
The number of work items (stories, tasks, or tickets) completed per sprint.
The cycle time for each item (how long it takes from start to finish).
The variability in completion rates over multiple sprints.
Even 10 sprints of data can provide useful insights.
Step 2: Pick a Monte Carlo Tool
While you can manually run Monte Carlo simulations in Excel, there are tools that automate the process:
Actionable Agile – Designed for Agile teams, with built-in Monte Carlo forecasting.
Kanbanize – Uses flow-based metrics to predict delivery timelines.
Jira Plugins (e.g., Forecast & Track) – Adds Monte Carlo analysis directly inside Jira.
Step 3: Run the Simulation
Once your data is ready, input it into your chosen Monte Carlo tool. The tool will:
Simulate thousands of possible future scenarios.
Provide a probability range for completion dates.
Offer confidence levels (e.g., “We are 90% confident this will be done in 12-14 sprints”).
Step 4: Use the Results for Smarter Planning
Now that you have probability-based forecasts, you can:
Set realistic expectations with stakeholders.
Adjust team capacity based on projected delivery timelines.
Identify risks early and adjust priorities accordingly.
Step 5: Continuously Improve
Monte Carlo forecasting isn’t a set-it-and-forget-it method. To get the best results:
Regularly update throughput data for better accuracy.
Compare actual outcomes to predictions and refine the model.
Educate your team and stakeholders on interpreting probability-based insights.
Conclusion: A Smarter, Data-Driven Approach to Agile Forecasting
Story points have been a fixture in Agile for years, but they come with significant downsides—they’re subjective, time-consuming, and don’t always lead to better predictions. Monte Carlo forecasting offers a better, data-driven alternative by using real historical throughput data to generate probability-based forecasts.
With Monte Carlo analysis, teams can make confident commitments, improve predictability, and remove the overhead of traditional estimation. More importantly, they can spend less time debating complexity and more time delivering value.
So, if you’re tired of endless estimation discussions, why not give Monte Carlo forecasting a try?
References
Vacanti, Daniel. When Will It Be Done?: Lean-Agile Forecasting to Answer Your Customers' Most Important Question. Leanpub, 2019.
Kersten, Mik. Project to Product: How to Survive and Thrive in the Age of Digital Disruption with the Flow Framework. IT Revolution, 2018.
Google. Accelerate: The State of DevOps Report. Retrieved from https://cloud.google.com/devops/state-of-devops
Rother, Mike. Toyota Kata: Managing People for Improvement, Adaptiveness, and Superior Results. McGraw-Hill, 2009.