> For the complete documentation index, see [llms.txt](https://mixmodeler.gitbook.io/mixmodeler-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://mixmodeler.gitbook.io/mixmodeler-docs/getting-started/introduction-to-mixmodeler.md).

# Introduction to MixModeler

### Welcome to MixModeler

MixModeler is a comprehensive, **privacy-first** Marketing Mix Modeling (MMM) platform that enables marketers and analysts to build sophisticated attribution models **without coding**. Our platform combines the statistical rigor of traditional econometric modeling with the accessibility of modern no-code interfaces.

#### What Makes MixModeler Different?

**Privacy-First Architecture**

* Your data **never leaves your device** - all processing happens locally in your browser
* No cloud storage of sensitive marketing data
* Perfect for enterprises with strict data governance policies
* Complete control over your information

**Professional-Grade Analytics**

* Both **Bayesian** and **OLS (Frequentist)** modeling approaches
* Advanced saturation curves and adstock transformations
* Comprehensive diagnostic testing suite
* Full uncertainty quantification with Bayesian methods

**No-Code Accessibility**

* Build complex MMM models through intuitive point-and-click interface
* Real-time preview of model changes before applying
* Interactive visualizations and charts
* Excel integration for seamless data import/export

**Enterprise Features at Accessible Pricing**

* **95% less expensive** than traditional MMM solutions
* Granger causality testing for variable validation
* Multicollinearity detection (VIF analysis)
* MCMC convergence diagnostics for Bayesian models
* Comprehensive model comparison tools

***

### What is Marketing Mix Modeling?

Marketing Mix Modeling (MMM) is a **statistical analysis technique** that quantifies the impact of various marketing activities on business outcomes like sales, revenue, or conversions. Unlike digital attribution which relies on cookies and user tracking, MMM uses **aggregate time-series data** to understand marketing effectiveness.

#### Key Questions MMM Answers

**Attribution & Impact**

* Which marketing channels drive the most incremental sales?
* What's the true contribution of each channel to business results?
* How do different channels work together (synergy effects)?

**Optimization & Planning**

* How should I allocate my marketing budget across channels?
* What happens if I increase/decrease spend on specific channels?
* Which channels show diminishing returns at current spend levels?

**ROI & Performance**

* What's the return on investment for each marketing activity?
* Which channels are most cost-effective?
* How does marketing performance vary over time?

#### Why MMM Matters Now More Than Ever

**Privacy Regulations**

* GDPR, CCPA, and other privacy laws limit tracking-based attribution
* Third-party cookie deprecation makes digital attribution unreliable
* MMM works with aggregate data - no personal information needed

**Complete Marketing View**

* Measures **ALL channels** including TV, radio, print, outdoor, sponsorships
* Captures online AND offline marketing impact
* Includes factors like seasonality, pricing, competitor activity

**Strategic Decision Making**

* Provides forward-looking insights for budget planning
* Enables scenario analysis ("what-if" testing)
* Supports data-driven marketing strategy

***

### Core Capabilities

#### Data Management

**Flexible Import & Processing**

* Excel file import with automatic data validation
* Support for .xlsx, .xls, and .CSV formats
* Data quality checks and preprocessing tools
* Handle weekly, monthly, or custom time periods
* **Subscription-based limits**: Free (20 vars), Professional (500 vars), Business (unlimited)

#### Variable Engineering

**Transform Your Data**

* **Lead/Lag variables**: Capture delayed effects
* **Saturation curves**: Model diminishing returns (S-shape and concave)
* **Adstock transformation**: Represent carryover effects (20-90% decay)
* **Weighted combinations**: Create composite metrics
* **AVO (Average Value Optimization)**: Smooth volatile data
* **Date splitting**: Isolate campaign or seasonal periods

#### Model Building

**Flexible Modeling Approaches**

* **OLS (Ordinary Least Squares)**: Fast, deterministic point estimates
* **Bayesian inference**: Full uncertainty quantification with credible intervals
* **Prior configuration**: Incorporate business knowledge into Bayesian models
* **Multiple model comparison**: Test different variable combinations
* **Real-time coefficient updates**: See changes immediately

#### Advanced Testing

**Validate Your Models**

* **Variable testing**: Pre-screen variables before adding to models
* **Granger causality**: Test if marketing truly predicts KPI changes
* **Multicollinearity detection**: Identify correlated variables (VIF scores)
* **Stationarity tests**: Ensure data is suitable for MMM analysis
* **Optimal adstock rates**: Find best carryover parameters for each channel

#### Diagnostic Analysis

**Ensure Model Quality**

* **Residual normality tests**: Jarque-Bera, Shapiro-Wilk
* **Autocorrelation detection**: Durbin-Watson, Ljung-Box tests
* **Heteroscedasticity tests**: Breusch-Pagan, White tests
* **Influential points analysis**: Cook's Distance, leverage metrics
* **Model fit metrics**: R-squared, Adjusted R-squared, AIC, BIC
* **Bayesian diagnostics**: R-hat, effective sample size, divergences

#### Decomposition & Insights

**Understand Channel Contributions**

* **Group-based decomposition**: Organize channels by business function
* **Time-series contribution**: See how each channel performs over time
* **Variable-level drilldown**: Analyze individual variables within groups
* **OLS and Bayesian decomposition**: Point estimates or uncertainty ranges
* **Visual attribution**: Interactive stacked charts with customizable colors

#### Visualization & Reporting

**Share Results**

* Interactive charts: time series, scatter plots, correlation heatmaps
* Publication-ready visualizations
* **Excel export** with complete model details, coefficients, and decomposition
* PDF diagnostic reports for stakeholder sharing
* Customizable contribution group colors

***

### Who Should Use MixModeler?

#### Marketing Analysts

**Build Professional Attribution Models**

* Optimize marketing spend allocation across channels
* Validate channel performance with statistical rigor
* Create data-driven budget recommendations
* Generate executive-ready reports and visualizations

#### Data Scientists

**Leverage Advanced Methods**

* Access both Bayesian and Frequentist approaches
* Configure priors and MCMC settings for Bayesian models
* Run comprehensive diagnostic tests
* Handle complex variable transformations and interactions
* Work with large datasets (up to unlimited variables on Business plan)

#### Marketing Managers

**Get Actionable Insights Without Technical Expertise**

* No coding or statistics PhD required
* Understand which channels drive results
* Make data-informed budget allocation decisions
* Justify marketing investments to leadership
* Track performance over time

#### Consultants & Agencies

**Deliver Professional MMM Analysis**

* Serve multiple clients efficiently
* Provide comprehensive reporting and documentation
* Maintain data privacy for sensitive client information
* Create reproducible analysis workflows
* Export results for client presentations

***

### Getting Started

Ready to build your first Marketing Mix Model? Here's your roadmap:

#### 1. **Prepare Your Data**

Gather time-series data with your KPI and marketing activities (minimum 26 weeks, 52+ recommended)

#### 2. **Create Account**

Sign up for MixModeler and choose your subscription plan based on dataset size

#### 3. **Upload Data**

Import your Excel file following our data format requirements

#### 4. **Build Models**

Follow our guided workflow to create your first MMM model

#### 5. **Analyze Results**

Run diagnostics, decomposition, and extract actionable insights

***

### Platform Architecture

#### Privacy-First Processing

**All computation happens in your browser:**

* Data never transmitted to servers
* No cloud storage or external databases
* Processing powered by WebAssembly (WASM) and optional GPU acceleration
* Suitable for highly sensitive marketing data

#### Performance Features

**Fast, Modern Technology Stack:**

* **GPU acceleration** for large datasets (if available)
* **WASM processing** for browser-based computation
* React-based responsive interface
* Real-time model updates and previews

#### Security & Compliance

**Enterprise-Grade Data Protection:**

* On-device processing eliminates data transfer risks
* No cookies or tracking of user behavior
* GDPR and CCPA compliant by design
* Ideal for regulated industries (finance, healthcare, etc.)

***

### Next Steps

Explore the detailed guides to master MixModeler:

* **Quick Start Guide**: Build your first model in 15 minutes
* **Data Upload & Formats**: Prepare your data correctly
* **Core MMM Theory**: Understand the methodology
* **Variable Engineering**: Transform variables for optimal performance
* **Model Building**: Create and refine your models
* **Decomposition Analysis**: Extract actionable insights

Let's dive into the core methodology that powers MixModeler's advanced analytics capabilities.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://mixmodeler.gitbook.io/mixmodeler-docs/getting-started/introduction-to-mixmodeler.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.