Budget Allocation — Performance Marketing Automation Workflow

schema_version: 1
version: "1.0.0"
id: budget_allocator
title: "Budget Allocation"
description: "Recommend optimal budget allocation across channels based on historical performance, efficiency metrics, and business constraints."
category:
  level: task
  domain: marketing
  reusable: true
params:
  total_budget: string
  channels: string
  historical_performance: string
  business_goal: string
  constraints: string
steps:
  - type: llm.generate
    prompt: |
      You are a performance marketing strategist optimizing budget allocation.

      **Total Budget:** {{total_budget}}
      **Active Channels:** {{channels}}
      **Historical Performance:** {{historical_performance}}
      **Business Goal:** {{business_goal}}
      **Constraints:** {{constraints}}

      Analyze the data and recommend budget allocation using this framework:

      1. **Efficiency Analysis**
         - Calculate CPA/ROAS efficiency score per channel
         - Identify channels at diminishing returns (rising marginal CPA)
         - Identify channels with headroom (below-target CPA, impression share available)

      2. **Allocation Method**
         - Rank channels by marginal efficiency
         - Allocate minimum viable spend per channel (floor for data sufficiency)
         - Distribute remaining budget proportional to efficiency
         - Reserve 10-15% for testing (new channels, campaigns, audiences)
         - Apply any hard constraints from the brief

      3. **Scenario Modeling**
         - Scenario A: Efficiency-optimized (maximize conversions at target CPA)
         - Scenario B: Growth-optimized (maximize volume, accept higher CPA)
         - Scenario C: Balanced (compromise between efficiency and volume)

      **Output format:**
      ## Current State
      [Table: Channel | Spend | Conversions | CPA | ROAS | Efficiency Score]

      ## Recommended Allocation
      [Table: Channel | Current $ | Recommended $ | Change % | Expected Impact]

      ## Scenarios
      ### Scenario A: Efficiency-Optimized
      [Allocation + projected outcomes]

      ### Scenario B: Growth-Optimized
      [Allocation + projected outcomes]

      ### Scenario C: Balanced
      [Allocation + projected outcomes]

      ## Recommendation
      [Which scenario to implement and why]

      ## Testing Budget
      [How to allocate the 10-15% test budget]

      ## Review Cadence
      [When to reassess this allocation]
    as: allocation

  - type: variable.set
    name: result
    value: "{{allocation}}"