> ## Documentation Index
> Fetch the complete documentation index at: https://help.superhawk.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Metrics

> The product metrics Superhawk tracks for your accounts, and how to configure them.

Metrics are how Superhawk measures what your accounts actually do in your product. They are the quantitative half of every health read: "contract creation down 49% quarter on quarter" is a metric, computed per account, on a schedule.

Metric configuration lives under **Configure Metric** in CS and hybrid workspaces.

## What a metric is

A metric has:

* **A name** in your product's language -- "Contracts created", "Active signers", "Templates in use"
* **A source** -- your data warehouse, product analytics, or another connected system
* **A window** -- the period it is computed over, such as the trailing 30 or 90 days
* **A scope** -- computed per account, so every account gets its own value and its own trend

Once defined, a metric is computed on a schedule for every account and becomes available to health judgements, signals, briefs, and reports.

## Configuring a metric

<Steps>
  <Step title="Open Configure Metric">
    From the sidebar, click **Configure Metric**. You will see the metrics already defined for your workspace.
  </Step>

  <Step title="Choose or create a metric">
    Open an existing metric to adjust it, or create a new one.
  </Step>

  <Step title="Define what it measures">
    Give it a name, point it at its source, and define how the value is computed for a single account.
  </Step>

  <Step title="Set the window">
    Choose the period the metric is computed over. This is the most consequential setting on the form -- see the warning below.
  </Step>

  <Step title="Preview before saving">
    Run a preview against real accounts and check the numbers against something you already know to be true. A metric that computes cleanly and means the wrong thing is worse than one that fails loudly.
  </Step>
</Steps>

<Warning>
  **Always check the window explicitly.** A metric left on the default window will compute successfully, produce plausible-looking numbers, and quietly measure a different period from the one you intended. It generates no error, so the only way to catch it is to look.
</Warning>

## Where metrics show up

* **Account health** -- as the factors behind a band, with the number quoted
* **Feature usage** -- as the volume and trend behind a feature
* **Signals** -- as the trigger for a usage-based observation
* **Briefs and reports** -- as the numbers you take into a customer conversation

## Getting metrics right

**Measure the thing the customer buys.** A metric on the workflow the customer is paying to accelerate is worth ten metrics on logins.

**Prefer a metric with a natural direction.** "Contracts created" is readable in isolation. "Average page depth" needs a paragraph of explanation before anyone can act on it.

**Watch the trend, not the level.** Large customers have large numbers. The question is whether this quarter is better or worse than last.

## Related

* [Features & Modules](/customer-success/modules) -- the vocabulary metrics are read in
* [Account Health](/customer-success/account-health) -- where metrics land
* [Data Warehouse](/integrations/warehouse/bigquery) -- a common metric source