Marketing KPI Dashboard AI Prompt

Spec a marketing KPI dashboard with this AI prompt. Channel ROI, CAC, and attribution metrics that prove marketing works. Free to copy and use today.

The prompt

ROLE: You are a marketing analytics lead who builds dashboards that survive CFO scrutiny.

CONTEXT: Business: [BUSINESS]. Channels: [CHANNELS]. Marketing budget: [BUDGET]. Funnel stages: [FUNNEL]. Attribution approach: [ATTRIBUTION]. Key question: [QUESTION].

TASK: Design a marketing KPI dashboard spec.

FORMAT: 1. Sections: Spend and Budget (by channel, pacing vs plan), Funnel Performance (stage volumes and conversion rates), Channel ROI (CAC, payback, ROAS per channel), Content/SEO (if relevant: traffic, rankings, assisted conversions), Brand (share of voice, direct traffic trend). 2. Per metric: definition, target, visualization, data source. 3. CAC Calculation Standard (define exactly what goes into CAC so finance agrees). 4. Monthly Review Agenda (30 min: what to look at in order). End with 3 attribution pitfalls and how this dashboard avoids them.

CONSTRAINTS: CAC definition must be agreed with finance before building; it is the #1 dashboard argument. Show incrementality thinking, not just last-click. Every channel metric needs a spend context; ROI without spend is meaningless.

Variables to fill in

Replace each [variable] in the prompt with your own details:

[BUSINESS] [CHANNELS] [BUDGET] [FUNNEL] [ATTRIBUTION] [QUESTION]

Example input

BUSINESS=B2B SaaS $50k ACV, CHANNELS=paid search, LinkedIn, content/SEO, events; BUDGET=$120k/month; FUNNEL=MQL, SQL, opportunity, closed; ATTRIBUTION=first-touch + self-reported; QUESTION=which channels truly drive pipeline?

How to use this prompt

Lock the CAC definition with finance first, then build; the dashboard's credibility depends on that single agreement.

Tips for the best results

Add self-reported attribution ('how did you hear about us'); it catches what tracking misses, especially dark social.
Review channel payback, not just CAC; a cheap channel with 18-month payback can be worse than a pricey fast one.
Hold out 10% of budget for experiments; dashboards optimize the known, experiments discover the new.