Unified Marketing Platform: 4-Wk Pilot Implementation

Tredence Inc

One stop solution for generating marketing ROI readouts to fulfill targeting, measurement & optimization needs of the marketing function

Objective: Optimize marketing decision making through a Unified Marketing Platform powered by a robust marketing data lake and robust ML models

Key Challenges Addressed:

  1. Fragmented view of customer/marketing data across legacy systems
  2. Inefficient data models leading to high dependency on IT/technical team
  3. Lack of customer/marketing insights to drive key investment decisions
  4. Scope of data consumption layer is limited

How do we address your challenges: 1.The acceleration layer uses 15+ accelerators to develop one analytics workbench for customer science, insights & measurement, C-journeys etc. 2. Multiple data consumption layers available for multiple personas including built in dashboards, ad-hoc query simulator etc. 3. Customer is the center of every decision with robust AI/ML analytics and visualizations 4. Integration of Marketing, Finance and Planning to enable faster decision making

Pilot Outcome: 1 brand; 1 region

Data harmonization bringing together marketing spends and performance data Initial version of ROI readouts through a standardized media mix model

Implementation Plan The break-up of the implementation plan is as below: Week 1 - Finalize the data elements inventory and map it to best data source Week 2 - Harmonize marketing spends and performance data from omnichannel investments Week 3-4 - ML models for evaluating impact of marketing spends by channel and generate marketing KPIs

This implementation uses the following native Azure components:

  1. Azure Data Factory: For data extract-transform-load (ETL) to deal with data coming from disparate internal and external systems.
  2. Azure App Services and Power BI: For developing web apps and dashboards for users to consume insights
  3. Azure Databricks: For developing ML models at scale
  4. Azure analysis services: To develop and maintain data models
  5. Azure SQL Database: To store ML model output for further consumption through dashboards or simulation