📊 Ad Performance Pipeline

Automated, end‑to‑end pipeline for Disney theatrical ad campaigns that consolidates Prisma, Innovid, and GCM360 data into clean KPI outputs for reporting and dashboards.

🎯 Objective

Automate ingestion and transformation of cross‑platform campaign data to deliver reliable KPIs with one command.

❌ Problem

Manual weekly reporting across CSV exports was slow and error‑prone, causing inconsistent metrics and delays.

✅ Solution / Approach

• Ingest: Read standardized CSVs for Prisma, Innovid, and GCM360 (see /data).
• Transform: Merge on (date, campaign_id), clean fields, compute KPIs (CTR, CPM, VCR, Pacing vs Plan).
• Deliver: Export outputs/consolidated.csv and outputs/report.xlsx with Daily KPI and Summary KPI tabs.

🔧 How It Was Solved

• Python ETL using pandas, clear merge logic, and KPI calculations.
• Minimal dependencies (pandas, openpyxl) for easy setup.
• Simple, reproducible structure suitable for recruiters and teams.

🔄 Flowchart

flowchart TD
    A["CSV inputs (prisma.csv / innovid.csv / gcm360.csv)"] --> B["ETL - pandas"]
    B --> C["Clean & merge"]
    C --> D["Compute KPIs (CTR CPM VCR Pacing)"]
    D --> E["Outputs (consolidated.csv & report.xlsx)"]

Loading

🚀 Quickstart

# 1) Create a virtual environment (optional but recommended)
python -m venv .venv && . .venv/Scripts/activate  # Windows
# or
python -m venv .venv && source .venv/bin/activate  # macOS/Linux

# 2) Install requirements
pip install -r requirements.txt

# 3) Run the pipeline
python src/pipeline.py

# 4) See the results
#   - outputs/consolidated.csv
#   - outputs/report.xlsx  (tabs: Daily KPI, Summary KPI)

📂 Project Structure

Ad_Performance_Pipeline_Full/
├─ data/            # Ready-to-run sample data
├─ outputs/         # Pipeline outputs (created on run)
├─ src/             # ETL code
├─ notebooks/       # (Optional) scratch space
├─ requirements.txt
└─ README.md

📈 KPIs Produced

• CTR = clicks / impressions
• CPM = cost / impressions * 1000
• VCR = video_completes / video_starts
• Pacing vs Plan = (cost − planned_spend) / planned_spend

Purpose‑built for Disney theatrical campaigns. Run once to generate outputs with no extra-setup.