Metrics

Metrics are essential for understanding how feature rollouts impact user behavior, helping to identify the optimal settings for your flags. Created from Events, they allow you to measure and aggregate various aspects of your application, such as page load times, conversion rates, or purchase amounts.

Create metrics that reflect your business objectives, and link them to feature flags within Experiments, enabling you to track how flag configurations influence outcomes over time.

View Metrics

View your Workspace's metrics by navigating to Metrics Dashboard.

Create a Metric

To create a metric, click the "New Metric" button from the Metrics Dashboard.

1. Enter a metric name, such as "Average Purchase Values per User"

2. Optionally add a description to explain the metric.

3. Select an Event, which has been submitted from the SDK with the track function.

Unlike other feature management platforms, events submitted to FlagSync from the SDK will automatically appear in the "Select Event" dropdown when creating a metric. There is no need to first create an Event and use that key when calling track.

This allows Product and Developers to track whatever events they feel may be useful in the future without first defining them in advance.

4. Select the value to aggregate from the Event, or use the default value.

See the track function for more details on the event value.

5. Select your Measurement Type:

Conversion Rate

Measure the percentage of users who perform an action event.

The denominator is flag impressions, while the numerator is the number of unique users who triggered the specified event.

Numeric

Numeric metrics are calculated per user, meaning each individual user's events are calculated and added to a single data point, ensuring equal weighting.

6. If Numeric is chosen, select the metric aggregation method.

7. Select the desired impact of this metric:

Increase

This metric should increase when compared to the baseline variant.

Decrease

This metric should decrease when compared to the baseline variant.

What is the baseline variant?

When you create an Experiment from this metric, you'll choose a baseline variant from the linked feature flag.

The baseline variant is the control group that you'll compare the other experiment variants against, and is often the original version of whatever you're testing.

This allows you to compare the performance of one or more new variants against this established standard to determine if they offer any improvements.