/dm:simulate

Purpose

Run Monte Carlo simulation of marketing scenarios to predict revenue outcomes with probability distributions. Test channel mix changes, budget reallocations, new channel launches, and spending adjustments before committing real budget. This command models uncertainty explicitly — instead of single-point forecasts that hide risk, it generates thousands of simulated outcomes per scenario to show the full range of what could happen, with calibrated confidence intervals. Use it when the stakes are high enough that "expected ROI" alone isn't sufficient and you need to understand downside risk, upside potential, and the probability of hitting specific revenue targets.

Input Required

The user must provide (or will be prompted for):

• Scenarios to simulate: One or more marketing scenarios to model — each defined by a set of channel budgets and assumptions. A scenario might be "shift 30% of paid search budget to TikTok" or "launch YouTube Ads at $15K/month while maintaining current spend" or "cut display by 50% and redistribute to email and SEO." Each scenario must include channel-level budget allocations and can optionally include custom ROI assumptions per channel
• Channel parameters per scenario: For each channel in each scenario: monthly budget allocation, expected ROI with mean and standard deviation (e.g., "3.2x +/- 0.8x" for a channel with historical variance), and saturation point if known (the spend level beyond which returns diminish sharply). If the user doesn't provide standard deviations, estimate from historical brand data or industry benchmarks
• Projection period: Number of months to simulate forward — typically 3, 6, or 12 months. Longer projections carry wider confidence intervals due to compounding uncertainty
• Revenue target (optional): A specific revenue figure the user wants to evaluate probability of achieving — e.g., "What's the probability we hit $2M in Q3?" The simulation will calculate the exact probability of reaching this target per scenario
• Number of simulations (optional): How many Monte Carlo iterations to run per scenario — defaults to 10,000 which balances statistical precision with speed. Can increase to 50,000+ for high-stakes decisions where tighter confidence intervals matter
• Constraints (optional): Minimum or maximum spend per channel, total budget cap, or required channel presence — the simulation respects these constraints when modeling outcomes

Process