“We’ll deliver these 30 items by November 19th!”

What’s the problem with this statement? There are two major issues with it:

Firstly, it suggests that there is an actual obligation to deliver all of the 30 items by November 19th. And secondly, it implies that these exact 30 items will be finished by then.

Can we make that promise knowing that knowledge work is notorious for its unpredictable nature?

We should always remember that there is uncertainty involved and we should communicate that there is more than one possible scenario that might happen in the future.

This is where the concept of probabilistic forecasting comes into play.

Here is what it looks like:

“We can deliver 30 items by November 19th with an 85% probability of hitting that target.”

Pay attention to the wording here. We communicate the risk associated with a certain outcome. Is there a chance we’ll finish earlier? Absolutely! But it won’t take us longer than November 19th, and there is an 85% chance we’ll keep our commitment.

We don’t provide a single, certain delivery date as an answer; all we do is mitigate the risk of failure.

4 Key Strategies for Making Reliable Delivery Commitments

Now that we understand the concept of probabilistic forecasting, how can we use it to make reliable delivery commitments?

There are four main strategies that can help you come up with accurate results. In fact, there are tools at your disposal that will give you the answers you’re looking for in less than a minute!

Let’s explore how this works.

How to Forecast When a Single Item Will Be Finished

When it comes to making a reliable delivery commitment for a single work item, we use the Cycle Time Scatterplot. This chart visualizes all your completed tasks as dots scattered on a plot. Each task comes with its finished date and the time it has taken to complete.

Cycle Time Scatterplot by Nave | Image

Use the Cycle Time Scatterplot by Nave to accurately forecast your task delivery times. See a dashboard with your data now

The horizontal dotted lines stretching across the graph are called percentile lines. We use percentiles to understand how much time we need to finish our work.

In the example above, the 50th percentile on our Scatterplot points to 8 days. This means that half of the tasks so far have taken up to 8 days to be finished.

Now, we can say that there is a 50% chance of finishing any work item in less than 8 days. We also know that there is an 85% chance of delivering any item in up to 18 days.

What we are saying is that we commit to delivering the work in less than 18 days. It will probably take less than that, but it won’t take more than 8 days to be finished, and there is an 85% chance that we will hit that target.

How to Forecast When a Project Will Be Finished

The tool to make reliable probabilistic forecasts for your project (without spending any time and effort!) is called the Monte Carlo simulation. The simulation uses a computational algorithm to provide you with a range of possible delivery dates and the likelihood of reaching each of those delivery dates.

Monte Carlo Simulation by Nave | Image

Agile teams across the globe use Monte Carlo simulations as an alternative approach to making project forecasts. Try it now, it’s free for 14 days

Let’s say the scope of your project is 30 tasks and you want to start working on it on June 12th. The simulation will give you a range of possible dates for completing your project. In this example, there is an 85% likelihood of finishing your project by November 19th, compared to a 95% likelihood of finishing your project by Dec 6th.

Monte Carlo, unlike other approaches, will quantify the risk you’re managing in terms of percentages.

You can choose to take a higher risk to reach an earlier deadline or mitigate your risk by choosing a later date in the timeline.

The question is no longer “when will you be done?” The question now becomes “how much risk are you willing to take?”

How to Forecast How Much Work You Can Take in Your Next Sprint

Now, the majority of the Scrum teams we’ve been working with used to plan their spirit using hours or story point estimations. The problem with both of these approaches is that they are relative, there is intuition, judgment, and gut feeling that comes into play.

The concept of probabilistic forecasting uses the capacity of your team to help you evaluate how much work you can take in your next iteration.

The capacity of your team is measured by the rate at which they deliver work. To determine your capacity, look into your past performance data to evaluate how many items you have completed per sprint in the past 3 to 6 months. This is where the Throughput Histogram comes in handy.

Throughput Histogram by Nave | Image

Use the Throughput Histogram during your Sprint Planning meetings to determine how much work you can commit to. Get started for free

The Throughput Histogram shows the number of items you completed in a certain period.

Looking into the example above, we can see that in the past 6 months, there were 2 sprint in which this team managed to deliver 2 items, 5 sprints in which they finished 3 items, 1 sprint with 4 items, another sprint with 5 items, and so forth. You can now use this analysis in your next Sprint Planning to better understand your capacity.

This team can schedule at least 2 items in the next sprint. This is their absolute minimum. They guarantee they will deliver at least 2 items, and that commitment comes with a 98% certainty that they’ll hit that target.

There is a 70% chance that they will deliver at least 3 items. The chance that they will complete 4, 5, and even 6 items drops down to less than 30%. If your team commits to that number, I’d certainly question that decision.

How to Forecast How Much Work You Can Take in Your Next Release

When determining the amount of work to put into your next release, when we talk about long-term commitments, we use another variation of Monte Carlo simulations.

Monte Carlo Simulation 2 by Nave | Image

Probabilistic forecasting is among the most accurate methods for tackling release planning. See a dashboard with your data now, it’s free for 14 days

In this simulation, we set the release date as Sep 30th and stated that we want to start working on it on Jul 1st. The simulation tells us that there is an 85% probability of completing at least 16 work items during this time period.

The simulation is repeated tens of thousands of times before the results are presented in the form of a probability distribution with percentiles increasing from right to left. The further you go to the left, the greater the certainty of completing the desired outcome.

If you’re interested in learning all about the reliable and unreliable approaches to making future predictions, I couldn’t be more thrilled to welcome you to our Sustainable Predictability program.

Here is your action item: If you haven’t created your dashboard just yet, now is the time! Go ahead and connect your management tool with Nave, it’s free for 14 days, no strings attached

In your next planning meeting, use the strategies above to come up with a delivery commitment. When planning, decide how much risk you are willing to take. For complex work with many unknowns, use higher percentiles (e.g., 95th percentile). For simpler work, lower percentiles (e.g., 70th percentile) can be more suitable.

By all means, keep doing what you’re currently doing and introduce probabilistic forecasting in parallel. Then let people see, on their own with time, which approach is the fastest, easiest, and most reliable!

I wish you a productive day ahead and I’ll see you next week same time and place, for more managerial goodness! Bye for now.

Do you find this article valuable?
Rating: 5 stars (3 readers voted)