Compares blended ROAS to reported channel ROAS and highlights attribution distortion.

Before you run it, change these fields

• channel_roas: Use reported ROAS from ad platforms
• spend_by_channel: Must match the same period
• revenue_attribution_model: Last click, data-driven, platform reported

{
  "prompt_name": "Blended ROAS vs Channel ROAS Reality Check",
  "role": "You are a performance marketing auditor.",
  "task": "Compare blended ROAS to channel-level ROAS and identify attribution bias.",
  "inputs": {
    "currency": "GBP",
    "total_revenue": null,
    "total_marketing_spend": null,
    "revenue_attribution_model": "platform reported | GA4 last click | GA4 data-driven",
    "channels": [
      {
        "channel": "Google | Meta | TikTok | Affiliates | Influencers",
        "spend": null,
        "reported_roas": null
      }
    ]
  },
  "output_requirements": {
    "include_sections": [
      "Blended ROAS result",
      "Channel ROAS comparison table",
      "Where attribution inflation is likely happening",
      "Channels that benefit from demand capture",
      "Channels that are probably undervalued"
    ],
    "format": "Markdown",
    "constraints": [
      "Do not assume perfect attribution",
      "Call out inconsistencies clearly"
    ]
  }
}