Enhanced Data Governance with Metrics (BETA)

What are Metrics?

Metrics are a new feature for Sigma (currently in beta) that allow data analysts and engineers to add another layer of data governance to Sigma Datasets. Metrics are predefined aggregations that allow end users to create any visualization or grouping with the guarantee that someone from the analytics team has already validated the formula.

Here are some common examples of metrics,

  • Net Promoter Score
  • Retention Metrics
  • Return Rates
  • New Customer Acquisition

Each of the above metrics is measured differently across individual companies and even within those companies, so it can be vital to any analysis that teams are working off a single shared understanding.

Additionally, metrics are a great way to reduce redundant work. If everyone computes Return Rate in the same way, it makes sense to have a single shared definition everyone can leverage in their work.

Set Up

  1. Find a Dataset you want to add a metric to. Currently, metrics are set-up within Datasets on the Metrics tab.

  1. Select Edit to begin working with your dataset and adding metrics

  1. Select Add a Metric which brings you into a new interface where you’ll define the calculation and metadata for your metric.

  1. One you define your field, you can simply click “Publish” in the upper right and this will make your Metric available to anyone using your dataset.

Using a Metric

Metrics appear next to columns within the Explore section of Sigma. Typically, you’ll use metrics within visualizations to provide a quick analysis of whatever grouping you’re most interested in.

  1. Create a Visualization as a child element of the dataset where your metric is stored.
  2. Add a column to your x-axis like you would with any other visualization.
  3. Finally, move to the Metrics section and drag your metric into the “y-axis”

  1. Once you do this, Sigma will take the same calculation you added in the Metrics section of your dataset and plug it into the underlying query powering your new visualization.

Again, this ensures that your users are always using the correct calculations when creating visualizations directly from your datasets.

Some Caveats

For now, only child elements of datasets will accurately calculate Metrics. If you create a child element from a child element and try to reuse any Metrics you’ve created, the calculation won’t function at the appropriate grouping level. We’re actively working to improve this functionality.

For now, metrics cannot refer to other metrics when you’re defining them within datasets. This is functionality we’re looking to add over time.