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.
Server-side multi-touch attribution for Node.js and TypeScript. Track journeys, attribute conversions, measure real ROAS.
Enterprise B2B multi-touch attribution and budget optimization engine built with Polars, DuckDB, Pydantic v2, Gemini structured output, and Streamlit.
Server-side multi-touch attribution for PHP. Framework-agnostic core with Laravel and Symfony adapters.
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.
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.
Deterministic PHP engine for parsing observed acquisition context, applying first/last-touch rules, and building canonical events.
Server-side multi-touch attribution for Python. Flask middleware, framework-agnostic core, no ad-blocker blind spots
Enterprise Google Cloud BigQuery GA4 multi-touch attribution engine stacking raw event streams with 4-model revenue allocation and SHA-256 PII protection.
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
CDP composable et warehouse-native : résolution d'identité et attribution multi-touch sur un lakehouse ouvert (dbt, DuckDB, MinIO, Kestra, Streamlit).
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.
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.
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.
Multi-touch attribution modeling comparing Last-Click, First-Click, Linear, Time-Decay, and Position-Based models on a digital marketing dataset — quantifying how channel credit shifts across attribution rules.
Feature engineering for marketing mix modelling and multi-touch attribution in R. Geometric and Weibull adstock, four saturation curves, carryover selected by cross-validation against your KPI, and customer journeys built from raw event logs.
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.
To associate your repository with the multi-touch-attribution topic, visit your repo's landing page and select "manage topics."