We’re happy to announce that we’ve added a Plan Visualizer instrument only for Agile Mentors Neighborhood members. When doing agile planning, use this free instrument to forecast the probability that an agile or Scrum crew can full a certain quantity of labor inside a goal variety of iterations.
AMC member? Entry the Plan Visualizer instrument.
Pattern Agile Forecast
Suppose a crew has been requested to finish 240 story levels inside 8 iterations (marked with a pink dot within the picture beneath). The crew estimates that its future velocity will fall within the vary of 26–34. The numbers (26 and 34) are used to attract the strains exhibiting how a lot will be delivered.
A velocity of 26 for eight sprints will lead to 208 factors accomplished; the decrease line is beneath the goal of 240 indicated by the pink dot. The crew’s high-velocity estimate of 34 leads to 272 factors; that line is above the goal.
As a result of the specified scope falls between what’s achievable with the crew’s predicted high- and low- velocity estimates, the goal is possible. It’s not assured, however it’s achievable.
If the crew’s velocity vary estimates had been decrease, say 20–26, the pink dot would seem within the pink zone, which might point out that the specified quantity of labor will not be possible, even on the crew’s excessive velocity estimate.
Equally, if the low quantity within the estimated velocity vary is excessive sufficient to complete all the things, the pink dot might be within the inexperienced zone. This means the goal set of labor could be very achievable.
Speaking with Stakeholders
I exploit charts like this to reply questions on why sure requests are simply not doable.
A chart that exhibits a set of labor and an accompanying goal deadline within the pink zone makes it clear that, even on the crew’s optimistic velocity estimate, the specified work can’t be accomplished in time. (I do not trouble exhibiting the chart if the request lands within the inexperienced zone. We simply get to work and delight clients by ending early.)
What If the Forecast Exhibits We Can not End?
When the focused set of labor and deadline are above one or each velocity estimates, you may have a couple of choices:
- Prolong the deadline (improve the variety of sprints)
- Cut back the scope
- Prolong the deadline and scale back the scope
You should utilize the Plan Visualizer instrument to experiment with completely different situations. Improve the variety of sprints and see the way it seems to be. Cut back the scope and see the way it seems to be.
Use the instrument to collaborate with stakeholders and set the best expectations.
Forecasting Velocity
To find out the viability of a plan, you need to forecast velocity.
The Plan Visualizer means that you can manually enter a velocity vary. (You should utilize one other free instrument, the Velocity Vary Calculator, or a way known as resampling velocity to estimate the speed vary your self.) Or, in case you enter historic velocity numbers, the Plan Visualizer will forecast a spread for you.
(Should you don’t have any historic information, you possibly can nonetheless forecast velocity. This Superior Subjects in Agile Planning presentation exhibits a method to do this. Estimate Velocity as an Agile Marketing consultant exhibits one other)
Let’s check out how the Plan Visualizer forecasts a velocity vary.
What to Think about When Forecasting Velocity
One of many stuff you’ll need to know is your crew’s imply velocity. To calculate it, look again at as a lot information as you suppose related; I’d contemplate a 12 months a very good higher restrict. A crew has probably modified sufficient in a 12 months that older information gained’t be very predictive. (Right here’s some recommendation on predicting velocity in case your crew adjustments continuously.)
To forecast a future velocity, you additionally want to think about how far forward you might be planning. Should you’re solely planning one dash forward, it could be dangerous to make use of the crew’s common. That’s, in any case, an common. The crew undoubtedly skilled some sprints higher than the common and a few worse.
The crew on this chart has a mean velocity of 29. However prior to now, their velocity has been as little as 18 and as excessive as 37. In the event that they had been requested to forecast their velocity within the subsequent dash, their trustworthy reply can be wherever from 18 to 37.
However suppose as an alternative they’re requested to forecast its velocity over the following 1,000 sprints. With that many sprints, the excessive and low velocities will steadiness out. The crew can safely forecast that their velocity might be 29, their common.
(I’m, in fact, ignoring the truth that the crew might be studying and bettering over 1,000 sprints. And that tremendous new applied sciences will emerge. Properly, and likewise that the majority of them will most likely retire—since 1,000 two-week sprints equals 38 years!)
In sum, to forecast velocity precisely it is best to use a spread. Should you’re forecasting just one dash forward, the estimate vary needs to be extensive. The vary can slim as you look additional forward.
Utilizing a Prediction Interval
The Plan Visualizer forecasts a future velocity vary utilizing a statistical idea generally known as a prediction interval, which is given by this system:
Which may look sophisticated, so let me break it down piece by piece.
We begin with the crew’s common, or imply, velocity. We then add and subtract some quantity to that, and that’s all the things to the best of the plus-minus image. Right here’s what that system means.
What’s a t-value?
The t-value is predicated on how assured we need to be within the plan (we’ll use 90%) and the variety of historic velocity values you may have. Should you had a statistics course prior to now, chances are you’ll bear in mind t-values from that.
The Plan Visualizer calculates the t-value for you. However you may also search for t-values on-line or use a spreadsheet perform.
What’s the s within the system?
The s within the system is the usual deviation of the crew’s velocities. It’s a measure of dispersion. It’s how unfold out a crew’s velocities are. Groups whose velocity bounces round between 28 and 32 could have a smaller commonplace deviation than these with a velocity vary between 20 and 40.
What are the h and f within the system?
Beneath the sq. root signal we now have 1 divided by h added to 1 divided by f. The variables h and f signify the variety of historic and future sprints.
If, for instance, this crew has 10 sprints’ value of information then h is 10. Should you’re predicting their velocity for 5 future iterations then f is 5.
On this case we’d add 1 over 10 to 1 over 5 to get the worth beneath the sq. root signal.
Prediction Interval Calculation in Motion
Earlier than working an instance prediction interval calculation, bear in mind you don’t must do any of this work: the Plan Visualizer will forecast velocity for you.
For our instance, I’ll use the speed graph proven above.
Step one is to calculate the imply and the usual deviation. That is trivial in Excel or Google Sheets. We simply use the common and stdev features.
Doing that, we get a imply of 29.4 and a normal deviation of 5.75.
To get the t-value we use the spreadsheet perform T.INV.2T. It takes two parameters. First, the worth .10, which signifies we’re calculating a 90% prediction interval. This implies there might be a 5% probability the longer term velocity common might be above our prediction interval and a 5% probability it is going to be beneath it. And a 90% probability that the common future velocity might be inside the vary we’re calculating.
The second parameter T.INV.2T wants is the variety of historic iterations you may have minus one. We’ve got ten iterations of information, so we use T.INV.2T(.10, 9). That leads to a worth of 1.83.
Let’s fill these numbers into our equation:
We additionally know h, the variety of historic iterations of information, is 10. And let’s assume this crew is being requested to look forward 5 sprints, which is the worth for f. Placing these values into the equation we get:
From right here, it’s straightforward to resolve the remainder of the equation:
This means that over the following 5 iterations, this crew’s velocity will be predicted at someplace between 29.4 plus and minus 5.68. Including and subtracting, we get that velocity might be from 23.72 to 35.08.
We need to do our agile estimating in entire numbers. Use your judgment on whether or not to spherical up or down for every estimate. I’d say 23.72 is shut sufficient to 24 that I’m going to spherical that up. The 35.08 is so near 35 that I’m positively rounding it down.
Should you felt compelled to be very conservative in your reply you would all the time spherical each down.
At this level, we will say there’s a 90% probability the crew’s future velocity might be between 24 and 35.
Forecast with Confidence
You should utilize it to:
- Set lifelike expectations with stakeholders—visually present what’s achievable given your agile crew’s velocity.
- Experiment with scope vs. timeline trade-offs—shortly regulate dash counts and scope to seek out viable options.
- Enhance planning accuracy—forecast future velocity utilizing statistical modeling, not simply intestine intuition.
Log in to start out planning with confidence.
Final replace: April ninth, 2025
