multi-touch-attribution

Data driven multi-touch attribution modeling with Markov chains

[Research] Transformer 기반 광고 기여도 모델 제안

In this research paper, we used Google and Facebook conversion lift studies to calibrate our Multi-Touch Attribution results from Google Ads Data Hub (ADH). We assessed the feasibility of these conversion lift calibrations and the impact of using conversion lift results in the calibration adjustment.

Enterprise B2B multi-touch attribution and budget optimization engine built with Polars, DuckDB, Pydantic v2, Gemini structured output, and Streamlit.

Open-source framework for estimating whether a brand is becoming more or less likely to be chosen.

A curated list of attribution, measurement, and marketing analytics resources. Open-source libraries, commercial platforms, research papers, datasets, and the people thinking hard about which marketing dollar caused which revenue dollar.

Server-side multi-touch attribution for Node.js and TypeScript. Track journeys, attribute conversions, measure real ROAS.

Server-side multi-touch attribution for PHP. Framework-agnostic core with Laravel and Symfony adapters.

Deterministic PHP engine for parsing observed acquisition context, applying first/last-touch rules, and building canonical events.

Multi-touch attribution for Shopify stores. First-click to purchase tracking via Theme App Extension and Remix admin.

Server-side multi-touch attribution for Ruby and Rails. Track customer journeys, attribute conversions, know what works.

Server-side multi-touch attribution for Python. Flask middleware, framework-agnostic core, no ad-blocker blind spots

CDP composable et warehouse-native : résolution d'identité et attribution multi-touch sur un lakehouse ouvert (dbt, DuckDB, MinIO, Kestra, Streamlit).

End-to-end Multi-Touch Attribution pipeline implementing 7 attribution models including simple rule-based approaches as well as Markov Chain and Shapley Value models. This also includes a full MTA vs MMM reconciliation framework.

Enterprise Google Cloud BigQuery GA4 multi-touch attribution engine stacking raw event streams with 4-model revenue allocation and SHA-256 PII protection.

The Attribution Modeling for ETL project offers a comprehensive suite of tools and methodologies for implementing various marketing attribution models, including first-touch, last-touch, linear, time decay, and U-shaped models. These models are essential for understanding the impact of different marketing channels on customer conversions.

Multi-agent system built in Claude Code — applies 5 attribution models to B2B pipeline data, identifies funnel bottlenecks, ranks channels by ROI, models budget scenarios, and generates an interactive CMO dashboard

Real-time probabilistic identity resolution engine for streaming platforms. Resolves multi-user attribution with 78% accuracy at <100ms latency. GDPR/CCPA compliant.

Omni-channel marketing attribution pipeline using Markov Chains & Shapley Value models. Dockerized Python application connecting to BigQuery, processing user journeys, and generating DuckDB/Parquet outputs for Grafana & Power BI visualization.

A B2B SaaS company runs campaigns across 10+ marketing channels (Organic Search, Paid Search, LinkedIn, Webinars, Referrals, etc.). Leads flow through a 10-stage funnel — from Website Visit to Opportunity Won — touching multiple channels along the way.