Logo Lanfrica

Derrick-Ryan-Giggs/safaricom-intelligence

Domaine:

socioeconomicdigital infrastructure

Type de record:

dataset
Créateur:
Der
Hôte:
Cloud ELT pipeline for Safaricom PLC public financial disclosures — M-PESA, Revenue Decomposition, Kenya vs Ethiopia (FY2014–FY2026) # Safaricom Intelligence A cloud ELT pipeline that converts Safaricom PLC's public financial disclosure PDFs into a versioned, queryable BigQuery dataset, with a dbt (Fusion) transformation layer and a 3-page Looker Studio dashboard built around three analytical lenses: M-PESA growth, revenue mix, and Kenya vs. Ethiopia. ## Dashboard **Open the live dashboard →** Three pages, built on `safaricom-intelligence.dbt_rygiggs_mart.*`, spanning FY2008 through FY2026: | Page | What it shows | |---|---| | **1. M-PESA Intelligence** | Revenue trend from launch (KES 0.37bn, FY08) to today, M-PESA as % of total service revenue, Fuliza (merchant overdraft) customer growth, active merchant base over time (Lipa na M-PESA + Pochi La Biashara combined) | | **2. Revenue Mix** | Stacked-area and 100%-stacked-column views of voice, data, messaging, M-PESA, and fixed/IoT as shares of total service revenue - makes visible the decade-long structural shift away from voice (which genuinely declined in absolute KES terms during FY20–21, a real COVID-era effect, not a data artifact) toward data and M-PESA | | **3. Kenya vs Ethiopia** | Dual-axis revenue/customer-base chart, EBIT margin comparison, an Ethiopia "path to profitability" chart, and capex comparison - the story of a J-curve market entry: Ethiopia's loss got *worse* before it got better (-5.1bn → -30.7bn → -59.6bn → -61.1bn at the trough → -30.1bn, nearly halving in the most recent year) | Data Freshness is set to 12 hours on all three data sources. Read on for how it's built. ## Table of Contents - Problem Description - Solution Overview - Architecture - Tech Stack - Data Sources - Project Structure - Data Model - Infrastructure as Code (Terraform) - Data Ingestion and Orchestration - Transformations (dbt Fusion / dbt Cloud) - Data Quality: What Broke and How It Was Found - Known Data Limitations - Reproducibility: How to Run - Author ## Problem Description Safaricom PLC (Kenya's largest telecom operator and the compa …