Metrics Library

The Metrics Library is a central catalog of your business metrics. Instead of redefining “revenue” or “active customers” in every report, you define each metric once — with its source, aggregation, filters, and format — and reuse it everywhere.

You can find the Metrics Library under Data → Metrics in the sidebar.

Creating a Metric

Metrics come in two types:

Type	How it’s built
Derived	Built from a single source — pick a table or connected app, a field, a measure, and optional filters
Composed	Built from a formula that combines other building blocks (e.g., `A / B` for a ratio like AOV or ROAS)

Source-based (derived) metrics

A derived metric points at one data source and aggregates a field. You can select the source in two modes:

Dataset — pick a physical BigQuery table and column directly
Connected app — pick a canonical event from a connected source; the definition resolves to the right physical table and column at compile time

Canonical-event definitions are resilient: because they reference the event and property by name rather than a hardcoded table, they keep working as your underlying schema evolves.

Formula-based (composed) metrics

A composed metric assigns a single-letter label (A, B, C, …) to each building block — an existing metric or a custom query — and combines them with a formula, for example:


A / B        # revenue per order
(A - B) / B  # period-over-period growth

Measures

Each building block aggregates its field with one of the following measures:

Measure	Description
Sum	Sum of values
Count	Count of rows
Average	Average of values
Minimum	Minimum value
Maximum	Maximum value
Distinct Count	Count of unique values
Median	Median of values
Running total	Cumulative sum of all values up to each date — useful for ledger- or stock-style metrics like cash balance or inventory on hand

You can also add filters to each block (equals, contains, greater than, between, and more) and choose a display format: number, currency, percentage, or multiplier.

The Metrics Table

The library page lists all metrics in a data table showing each metric’s name, live value (computed over the last 30 days), status, type, category, and last update. Values are calculated on demand from your warehouse, so the table doubles as a quick health check on your key numbers.

Metric Detail

Opening a metric shows its full definition and a “How it’s calculated” panel — an audit surface that renders the per-block definition and the exact BigQuery SQL the engine executes. The panel is generated with the same helpers the engine uses, so the displayed query can never drift from the one that actually runs.

Draft vs Active

Every metric has a status:

Status	Meaning
Draft	Being defined or reviewed — not yet available for use in reports
Active	Published and usable across the product
Archived	Retired from the active list but preserved

New metrics start as drafts. Activate a metric when its definition is ready for use in reports.

Metric Trees

The library also includes a metric trees view (Data → Metrics → Trees), which lays out how metrics relate to each other — composed metrics and the building blocks they depend on — as a visual tree. Use it to trace a top-line KPI down to its inputs.

MCP and CLI Access

MCP tools

When connected via the MCP Server, AI tools can manage metrics programmatically:

Tool	Action
`metrics_list`	List all custom metrics
`metrics_get`	Read a metric definition by ID
`metrics_search`	Search metrics by name or description
`metrics_create`	Create a new metric (derived or composed)
`metrics_update`	Modify an existing metric definition
`metrics_activate`	Promote a draft metric to active
`metrics_delete`	Remove a metric

CLI

The Vendo CLI exposes the same operations under vendo metrics:


vendo metrics list
vendo metrics get <metricId>
vendo metrics create
vendo metrics update <metricId>
vendo metrics activate <metricId>
vendo metrics delete <metricId>

Data Catalog — browse the tables and columns metrics are built on
Reports — where active metrics are put to work
MCP Server — programmatic metric management