The following is a step-by-step guide on how to create a Group Forecast in Portfolio Forecaster. You now have more flexibility when creating your forecast and have the ability to save them as well.
To create a new Group Forecast, select Create New from the Home Page or Add New from the Navigation Bar.
| Panel | ||||||||
|---|---|---|---|---|---|---|---|---|
| ||||||||
Missing Versions? They were not supported during our initial release, but we hope to add them back soon! |
The first section is about defining what you would like to forecast.
Start by giving your Group Forecast a name and a Confidence Level. It’s best to keep forecast names unique and concise.
The default Confidence Level is 85%, a conservative percentile with a small amount of risk.
If you have little room for risk, we suggest staying at this level or above, but if you’re open to taking a gamble, feel free to lower the confidence level.
Begin by choosing the Data Source for your parent issue; you can choose from a Board, Project, or Filter (if you select a filter, enter the JQL in the designated area).
Then select the Issue Type(s) you would like to forecast.
By default, all of the selected issue types from your chosen data source that are in an active category will be included in your forecast. You may override this behaviour with a JQL or manually select statuses.
The second section is about defining what constitutes “remaining work” for your forecasted items.
To make accurate forecasts, it's essential to understand how much work is still pending.
Currently, we support the parent field in identifying linked issues, and we're excited to share that more options are on the horizon.
Then, choose the Issue Type will be counted as work to be done. You can have multiple Issue Types.
We’ll only include the issue types you choose, and as always, we’ll steer clear of anything in a completed status category.
If you need to tweak something, feel free to use the toggle to redefine "completed." You can even provide a specific JQL query to expand your search beyond issue types.
In the third section, you will confirm the amount of historical data you would like to use for this forecast.
| Panel | ||||||||
|---|---|---|---|---|---|---|---|---|
| ||||||||
Ideally, use the work from the same project for your historical data, as this is most likely to be the most similar. Can’t do that? Then, use a project similar to the historical throughput for your forecast. |
Our standard timeframe is 90 days, but feel free to adjust it to better suit your requirements.
The max is 360 days.
If you find yourself without historical data or need to forecast using alternative data, simply toggle the override option. This will allow you to provide a JQL query that specifies exactly which data we should pull.
After you select
| Status | ||||
|---|---|---|---|---|
|
| Panel | ||||||||
|---|---|---|---|---|---|---|---|---|
| ||||||||
Want to know what will happen if something changes in the future? Don’t worry, What If? scenarios will be making their comeback soon! |
Problem? If you encounter the error message above, you don’t have enough finished items to forecast due date likelihoods and dates for each item.
You can still create the forecast and rerun it when you have at least ten items that match your historical data settings.