Custom metrics

A custom metric is anything you define yourself for the active organization. There are two kinds: Conversion Sum (a count or total of selected platform events) and Formula (an expression built from other metrics). Once you create one, it's available to pick when you build reports.

Where they live

The Custom Metrics section of the Metrics page ("Calculated fields and grouped conversions.") holds a table of everything you've created:

• Display name — the label that shows in reports (e.g. Profit).
• System name — the short identifier used inside formulas (e.g. profit). Derived from the display name, not typed.
• Type — Conversion Sum or Formula.
• Created — date.
• Action menu — Edit / Delete.

The Create Metric button opens the metric modal in create mode.

Choosing a type

The first thing the form asks for is the Type:

• Conversion Sum — for counts or totals of platform conversion events that don't fit into Purchases / Revenue / Lead. Example: "Demo requests" — total of three platform-specific demo-related conversions.
• Formula — for metrics built from other metrics. Example: profit = #revenue − #cost.

The rest of the form switches based on the type.

Naming

You give a metric one name — the display name, written the way you would say it: Cost per Lead, Profit, Demo requests.

The system name underneath it is derived from what you type — Cost per Lead becomes cost_per_lead — and is the identifier formulas use. It is filled in as you type and cannot be edited: it is kept unique across the organization for you, so two metrics can never collide over it. Picking a preset keeps that preset's own short name (fb_thruplay_rate rather than a slug of its label) until you edit the display name, at which point the system name follows the edit like any other.

Editing a metric later changes its display name only. The system name stays put, so formulas that reference it keep working.

Conversion Sum

Fields:

• Display name (required) — the label in reports, e.g. Demo requests. This is the only name you type; the system name is filled in for you (see Naming below).
• What to sum — choose between:
  • Count — counts each matching event once.
  • Value — sums the monetary value on each event (use this when the conversion carries a value you want to total).
• Conversions — the same picker as Conversion mapping, grouped by ad account, search-filterable. Tick the conversions to include.

Save, and the metric appears in the table and is available to pick in reports.

Formula

Fields:

• Display name (required) — the report label, and the only name you type.
• Formula — the expression, written in the formula editor described below.

The formula editor

The editor is a single-line input that:

• Accepts numbers, parentheses, and the operators +, −, ×, ÷.
• Highlights metric references as chips. Type # to open the metric list where the cursor is and pick one — the list is ranked, so the closest match to what you have typed is first. Clicking a metric in the list below the field inserts it too.
• Accepts a name typed by hand in any case: write revenue, Revenue or REVENUE and it resolves to the metric when you leave the field.
• Supports Ctrl+Z / Ctrl+Y (⌘Z / ⇧⌘Z on a Mac) inside the field, so an undo edits the formula rather than the page around it.
• Validates as you go — unknown metric references and malformed expressions are flagged before you can save.

Platform metrics in formulas

Formulas aren't limited to system and custom metrics. The # autocomplete also offers platform metrics — the native numbers Google Ads and Meta report for your ad accounts: impressions, clicks, spend, video views, engagements, and more. Two things to know:

• Currency: monetary platform metrics are converted into the report's currency, so a formula mixing spend with mapped revenue stays consistent.
• Granularity: some platform metrics exist only at the campaign level, not for individual creatives. A formula using them shows values in campaign-based reports but not in creative-level breakdowns.

Formula presets

Some numbers the platforms show in their own dashboards aren't exposed as ready-made metrics — they're computed from other metrics. Level ships presets for the common ones, so you can add them in one click instead of writing the formula yourself:

Pick a preset and Level fills in the name, display name, and formula — you can tweak it before saving. A preset is hidden if the organization doesn't have the platform metrics it needs (for example, Meta video presets won't show for a Google-only organization).

A few example formulas

Note there's no need to build plain ROAS, CPA, AOV, CTR, CPM, or CPL — Level computes those for you. Formulas are for the variations the built-ins don't cover.

What you can't reference

• Metrics that don't exist yet. Build prerequisites first.
• Loops. A formula can't reference a metric that eventually references this one — the editor blocks the save.
• Anything other than numbers and other metrics. No variables, no settings.

Editing and deleting

• Edit opens the same modal pre-filled. You can change the name, display name, and formula / conversion list. Reports already using the metric pick up the new definition the next time they load.
• Delete removes the metric. Reports that referenced it lose that column when they next load — edit those reports to add a different metric or recreate this one.

Permissions

Creating, editing, and deleting custom metrics requires the Manage Metrics & Conversions permission. Members without it see the table and can view definitions but can't change them.