Enabling predictions

A prediction for a period metric or a period KPI projects your historical data into a possible future, and provides you a visual representation of past, current, and potential trends. Knowing these trends can inform your decisions and actions before an eventual situation arises.

1. Configure the prediction time range.
   1. From the Filters and predictions tab, toggle Prediction to the on position:
      You can configure predictions on a period metric or a period KPI.
   2. Select a time range from the drop-down list, or enter a custom time range.
      A time range is the period in the future that you want your prediction to cover. You can extend the time range for a prediction to a maximum of half the number of historical data points, limited to 50. A chart with a prediction displays the supported number of data points, regardless of the end date that you define in your time range.

      Tip: For the best predictive accuracy, the time range you set must include a minimum of six data points from your data set and a maximum of 50 data points. If the time range contains more than 50 data points, Business Performance Center uses the 50 most recent data points to make a prediction.
   3. Optional: Set lower and upper limits for predictions.
      After you set upper and lower limits, the predicted values in your chart stay within the bounds of these limits. Defining these limits ensures that the predicted values stay within a range that makes sense in a business context.
      If, for example, you have historical data for declining sales, a lower limit of 0 ensures that all projected sales volume does not descend below 0.
   4. Select Done to return to the dashboard.
2. Turn on predictions.
   1. Go to a chart that has predictions configured.
   2. From the overflow menu , select Show prediction.