metrics

The metrics section maps data columns to named metrics used in your deck. Every metric displayed must be defined here first.

There are three ways to define a metric:

1. Basic — one-to-one column mapping
2. Filter — selective aggregation with a query
3. Function — derived from other metrics

Basic metric

metrics:
  Impressions:
    column: main_metrics.Impressions    # or just "Impressions" for CSV
    aggf: sum
    metric_comparison_method: %         # optional

Field	Required	Values	Description
column	Yes	String	Data column name. Namespaced as query_name.column for DB sources.
aggf	Yes	sum, mean, min, max, last	Aggregation function for daily → weekly/monthly rollup.
metric_comparison_method	No	%, bps	Format for auto-generated WOW/MOM/YOY comparisons. Default: %.

Choosing an aggregation function

Use this	When values...	Example metrics
sum	Accumulate over time	Revenue, page views, orders, clicks
mean	Are rates or averages	Defects/million, conversion rate, DAU
min	You want the lowest	Min response time
max	You want the highest	Peak concurrent users
last	You want the most recent	Inventory level, account balance

Auto-generated growth rates

For every metric, the app auto-generates three derivatives:

Suffix	Meaning
{MetricName}YOY	Year-over-year growth rate
{MetricName}MOM	Month-over-month growth rate
{MetricName}WOW	Week-over-week growth rate

You do not define these — just reference them directly in your deck:

deck:
  - block:
      ui_type: 6_12Graph
      title: Page View YOY Growth
      y_scaling: "##.1%"
      metrics:
        PageViewsYOY:
          line_style: primary
          graph_prior_year_flag: false

Filter metric

Selectively aggregate a column based on a condition.

metrics:
  USRevenue:
    filter:
      base_column: "RevenueUSD"
      query: "Country == 'US'"
    aggf: sum

Field	Description
base_column	Column to aggregate
query	Pandas-style query. Operators: ==, !=, >, <, >=, <=

Example: filtering by department

Given this data:

Date	Department	Applicants
15-Jan-2022	Engineering	45
15-Jan-2022	Sales	30

metrics:
  ApplicantsEngineering:
    filter:
      base_column: "Applicants"
      query: "Department == 'Engineering'"
    aggf: sum

  ApplicantsSales:
    filter:
      base_column: "Applicants"
      query: "Department == 'Sales'"
    aggf: sum

Function metric

Derive a new metric from other previously defined metrics.

metrics:
  ClickThruRate:
    metric_comparison_method: bps
    function:
      divide:
        - metric:
            name: Clicks
        - metric:
            name: Impressions

Available functions: sum, difference, divide, product.

Aggregation happens first

Each input metric is aggregated independently, then the function is applied. For example, divide divides the aggregated values, not row-by-row daily values.

Variance to plan example

metrics:
  PageViews:
    column: main_metrics.PageViews
    aggf: sum

  PageViews Target:
    column: main_metrics.PageViews__Target
    aggf: sum

  VarianceToPlan:
    function:
      difference:
        - metric:
            name: PageViews
        - metric:
            name: PageViews Target

  VarianceToPlanPct:
    function:
      divide:
        - metric:
            name: VarianceToPlan
        - metric:
            name: PageViews Target