--- name: media-mix-modeling-09fe8ef34c0c description: Advanced econometric modeling for marketing effectiveness and budget optimization allowed-tools: - Read - Write - Glob - Grep - Bash - WebFetch metadata: specialization: marketing domain: business category: Marketing Analytics skill-id: SK-019 graph: domains: [domain:marketing] skillAreas: [skill-area:brand-strategy, skill-area:brand-positioning, skill-area:content-marketing] workflows: [workflow:brand-campaign-launch] roles: [role:marketing-manager, role:marketing-strategist, role:brand-manager] --- # Media Mix Modeling Skill ## Overview The Media Mix Modeling Skill provides advanced econometric modeling capabilities for measuring marketing effectiveness and optimizing budget allocation. This skill enables marketing mix model development, channel contribution analysis, saturation curve modeling, and scenario planning using statistical techniques and machine learning approaches including Google Lightweight MMM and custom Python/R implementations. ## Capabilities ### Marketing Mix Model Development - Bayesian model specification - Frequentist regression modeling - Time series decomposition - Variable selection and feature engineering - Model training and validation - Holdout testing and backtesting - Model diagnostics and validation - Documentation and reproducibility ### Channel Contribution Analysis - Base vs. incremental decomposition - Channel-level contribution calculation - Marginal contribution analysis - Diminishing returns identification - Channel interaction effects - Year-over-year contribution comparison - Share of contribution trending - Contribution waterfall visualization ### Saturation Curve Modeling - Diminishing returns function fitting - Hill function parameterization - S-curve response modeling - Optimal spend level identification - Saturation point calculation - Response curve visualization - Per-channel curve comparison - Confidence interval estimation ### Adstock/Carryover Effects - Adstock decay estimation - Carryover rate calculation - Geometric decay modeling - Weibull decay functions - Peak lag identification - Effective frequency calculation - Media half-life analysis - Long-term effect quantification ### Budget Optimization Algorithms - Constrained optimization - Marginal ROI maximization - Budget allocation simulation - Channel mix optimization - Spend threshold identification - Diminishing returns avoidance - Cross-channel trade-off analysis - Multi-objective optimization ### Scenario Planning - What-if budget scenarios - Channel reallocation modeling - Seasonal budget planning - New channel introduction simulation - Budget cut impact analysis - Growth scenario modeling - Competitive response scenarios - Economic downturn planning ### Incremental Lift Calculation - Incremental revenue estimation - Lift over baseline calculation - Test vs. control comparison - Geo-matched market testing - Synthetic control methods - Causal impact analysis - Incrementality confidence intervals - Attribution vs. incrementality reconciliation ### Cross-Channel Synergy Analysis - Interaction term modeling - Synergy effect quantification - Complementary channel identification - Cannibalization detection - Optimal channel combination - Sequencing effect analysis - Cross-media amplification - Halo effect measurement ### Seasonality Adjustment - Seasonal pattern identification - Holiday effect modeling - Trend decomposition - Cyclical pattern adjustment - Weather impact incorporation - Event-based adjustment - Calendar normalization - Forecast seasonality application ## Process Integration This skill integrates with the following marketing processes: - **marketing-roi-analysis.js** - ROI calculation and budget optimization - **attribution-modeling-setup.js** - Attribution model calibration - **integrated-campaign-planning.js** - Budget allocation and planning ## Dependencies - Python data science libraries (pandas, numpy, scipy, statsmodels) - Google Lightweight MMM / Robyn - R statistical libraries - Bayesian modeling frameworks (PyMC, Stan) - Optimization libraries (scipy.optimize, cvxpy) - Visualization libraries (matplotlib, plotly) ## Usage ### Model Development ```yaml skill: media-mix-modeling action: build-model parameters: model_type: bayesian_mmm framework: lightweight_mmm data_configuration: date_column: week target_variable: revenue media_variables: - tv_spend - digital_display_spend - paid_search_spend - paid_social_spend - radio_spend control_variables: - price_index - competitor_spend - economic_indicator - seasonality_index model_settings: adstock: type: geometric max_lag: 8 saturation: type: hill priors: type: informative source: prior_mmm_results validation: holdout_weeks: 12 cross_validation_folds: 5 ``` ### Channel Contribution Analysis ```yaml skill: media-mix-modeling action: analyze-contributions parameters: model_id: "mmm_2024_q4" analysis_period: start_date: "2024-01-01" end_date: "2024-12-31" outputs: - type: contribution_breakdown format: waterfall_chart - type: channel_roi format: bar_chart - type: contribution_over_time format: stacked_area - type: marginal_contribution format: line_chart export: format: [pdf, csv, xlsx] destination: "reports/mmm_contributions" ``` ### Budget Optimization ```yaml skill: media-mix-modeling action: optimize-budget parameters: model_id: "mmm_2024_q4" optimization_settings: objective: maximize_revenue total_budget: 10000000 constraints: - channel: tv_spend min_percent: 0.20 max_percent: 0.40 - channel: paid_search_spend min_percent: 0.15 max_percent: 0.30 - channel: paid_social_spend min_percent: 0.10 max_percent: 0.25 business_rules: - type: minimum_presence channels: [tv, digital_display] - type: maximum_concentration single_channel_cap: 0.50 scenarios: - name: "optimal_allocation" constraints: default - name: "digital_first" overrides: digital_channels_min: 0.60 - name: "brand_building" overrides: tv_min: 0.35 ``` ### Scenario Planning ```yaml skill: media-mix-modeling action: run-scenarios parameters: model_id: "mmm_2024_q4" scenarios: - name: "Budget Cut 20%" budget_change: -0.20 allocation: optimized - name: "Budget Increase 30%" budget_change: 0.30 allocation: optimized - name: "TV Elimination" channel_changes: tv_spend: 0 reallocate: true - name: "New Channel Test" new_channels: - name: connected_tv estimated_roi: 2.5 test_budget: 500000 - name: "Q1 Seasonal Plan" period: "2025-01-01 to 2025-03-31" seasonality_adjustment: true comparison_metrics: - total_revenue - incremental_revenue - overall_roi - channel_roi ``` ### Saturation Analysis ```yaml skill: media-mix-modeling action: analyze-saturation parameters: model_id: "mmm_2024_q4" channels: - tv_spend - paid_search_spend - paid_social_spend analysis: - type: response_curves spend_range: [0, 2x_current] granularity: 100_points - type: optimal_spend threshold: 0.95_saturation - type: marginal_roi_curve spend_range: [0.5x_current, 1.5x_current] visualization: charts: - response_curves_overlay - marginal_roi_comparison - saturation_heatmap ``` ## Best Practices 1. **Data Quality**: Ensure sufficient historical data (2+ years recommended) 2. **Variable Selection**: Include relevant control variables (pricing, competition, economy) 3. **Model Validation**: Use holdout periods and cross-validation 4. **Uncertainty Quantification**: Report confidence intervals, not just point estimates 5. **Regular Refresh**: Update models quarterly with new data 6. **Triangulation**: Validate MMM results with experiments where possible 7. **Stakeholder Communication**: Present results in business-friendly formats 8. **Documentation**: Maintain model documentation and assumptions log ## Related Skills - SK-005: Marketing Analytics Platform - SK-014: BI and Dashboard Platform - SK-018: CRM Integration ## Related Agents - AG-008: Marketing Analytics Director - AG-012: Media Planning Expert --- Source attribution: a5c-ai Original project: https://github.com/a5c-ai/babysitter License: MIT License information: https://opensource.org/license/mit Original upstream instructions; downloaded from Dave AI. ## Upstream license notice MIT License Copyright (c) 2026 Gabriel Moreira Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. # Dave AI resource metadata Creator / publisher: a5c-ai Original project: https://github.com/a5c-ai/babysitter Attribution basis: Upstream repository explicitly recorded in package mirror path Fields of work: Data & Analytics, Marketing, Sales Roles: Business Development, Data Analyst, Growth Marketer, Marketing Analyst, Product Marketer Specialties: API, Database Tools: Python, R Country applicability: Country not specified Classification basis: Source catalog labels Instruction SHA-256: 4bb56f830d49a39e053f49db731ea0a590d3d583f4604a2551e089ff793be5d8