vs
Haus vs Recast
Haus and Recast are both Marketing Measurement (MMM, MTA, Attribution) tools. Here is how they compare on pricing, free plans, AI integrations and platforms, and when each is the better fit.
Side by side
| Feature | Haus | Recast |
|---|---|---|
| Category | Marketing Measurement (MMM, MTA, Attribution) | Marketing Measurement (MMM, MTA, Attribution) |
| Pricing | Quote-based | Quote-based |
| Free option | Not listed | Not listed |
| AI integrations | — | MCP server, Claude |
| Best for | Enterprise brands, Performance marketing teams | — |
| Links | Haus listingVisit website (opens in a new tab) | Recast listingVisit website (opens in a new tab) |
Which to choose
Choose Haus if
- want both geo-based and time-based holdout testing alongside media mix modelling
- need causal attribution that tracks lift down to individual campaign ad creatives
- want an internal AI agent (Architect) to flag budget risks and run scenario simulations
- sell on Shopify or Amazon and want built-in ecommerce platform integrations
Choose Recast if
- want Bayesian marketing mix modelling that tracks time-varying ROIs and multi-stage funnel interactions
- value out-of-sample validation scorecards to check predictive accuracy against past results
- want to query model outputs directly inside Claude, ChatGPT or Codex via Model Context Protocol (MCP)
- need to set finance targets and forecast revenue under custom channel caps and spend constraints
Verdict
Haus and Recast both combine incrementality experiments with marketing mix modelling for enterprise brands and performance teams. Haus suits marketers who want weekly causal MMM updates combined with multi-channel attribution down to the ad creative level. It also includes an internal AI agent, Architect, for scenario simulations and connections to Shopify and Amazon, though non-geo testing and advanced MMM sit on higher tiers.
Recast suits growth and finance-minded teams that prioritise Bayesian statistical modelling, GeoLift testing, and transparent out-of-sample validation scorecards. Rather than a strictly internal AI assistant, Recast offers an MCP server and Claude plugin so teams can query models in external tools like Claude and ChatGPT. The trade-off is that Recast requires custom onboarding setup and relies heavily on high-quality historical data.
FAQ
- How much do Haus and Recast cost?
Both Haus and Recast use quote-based pricing and do not publish standard rates publicly. Neither platform lists a free option, and Haus requires booking a demo for pricing details.
- How do Haus and Recast use AI differently?
Haus includes an internal AI agent called Architect that analyses experimental data, highlights budget risks, and runs what-if budget simulations. Recast uses the Model Context Protocol (MCP) and a Claude plugin to connect its model outputs and report templates directly into external AI assistants like Claude, ChatGPT, and Codex.
- What is the main difference in their measurement methodologies?
Haus combines geo-based and time-based holdout experiments with weekly causal MMM and creative-level attribution. Recast focuses on proprietary Bayesian MMM calibrated with location-based GeoLift tests, featuring out-of-sample validation scorecards and time-varying ROI tracking.
- Which tool is better for ecommerce brands on Shopify or Amazon?
Haus explicitly lists integrations with ecommerce platforms including Shopify and Amazon, alongside major ad networks and data warehouses. Recast does not specify native Shopify or Amazon connectors, focusing instead on public APIs and MCP connections.