DEX Score normalization for metric scores

  • Release version: Australia
  • Updated March 12, 2026
  • 1 minute to read

    • Learn how normalization adjusts individual scores of each metric to a common scale, enabling comparison across different metrics.

    In Digital Experience Score​, the key metrics monitored from applications and devices are normalized by mapping the Good, Average, or Poor range of the actual metric value to the corresponding Good, Average, or Poor range definition on a scale of 100.

    Normalization of metric values is done in two ways: One to maximize values and the other to minimize values.
    To maximize the value

    Maximize the value when higher values are better, for example, metric for WiFi reception rate. The following formula is used to get the normalized score:

    Normalized value (Maximize) = L1 + ((M1 - L2) / (H2 - L2) x (H1 - L1))
    To minimize the value

    Minimize the value when lower values are better, for example, metric for CPU usage. The following formula is used to get the normalized score:

    Normalized value (Minimize) = H1 - ((M1 - L2) / (H2 - L2) x (H1 - L1))
    Where,
    • L1 = Low value of mapping range
    • H1 = High value of mapping range
    • M1 = Actual metric value
    • L2 = Low value of metric value range
    • H2 = High value of metric value range
    You can customize the lower bound values for Good, Average, and Poor ranges. For more information, see Customize metric definitions
    Example for an application metric normalization (maximize)
    Consider the following values for the metric WiFi reception rate:
    • Actual metric value (M1): 400 (falls in good range)

    • Mapping range for good (L1 and H1): 75–100
    • Metric value range for average (L2 and H2): 50–500

    Then the application health metrics score is calculated as:

    Normalized value (Maximize) = 75 + ((400 - 50) / (500 - 50) x (100 - 75)) = 94 (rounded off).

    Example for a device metric normalization (minimize)
    Consider the following values for the metric CPU Usage:
    • Actual metric value: 79 (falls in average range)
    • Mapping range for good (L1 and H1): 65–74
    • Metric value range for average (L2 and H2): 55–84

    Then the device health metrics score is calculated as:

    Normalized value (Minimize) = 74 - ((79 - 55) / (84 - 55) x (74 - 65)) = 66 (rounded off).