PPC Performance Dashboard

An interactive paid-media performance dashboard built with Python, Pandas, Plotly, and Streamlit. Upload a Google Ads / Meta Ads / Bing Ads export, and get normalized KPIs, filters, trend charts, campaign and ad group breakdowns, period-over-period comparison, and CSV export — all in one page.

Problem it solves

Marketers waste hours stitching CSV exports into Excel to answer simple questions: "How did CTR change this month?", "Which ad group has the lowest cost per conversion?" This dashboard answers those questions instantly with consistent KPI definitions (CTR, CPC, CPA, conversion rate) and honest, clearly-labeled derived metrics.

Key features

• Upload CSV or XLSX exports; column names are normalized automatically (date, campaign, ad group, impressions, clicks, spend, conversions)
• Automatic KPI calculation: CTR, CPC, Cost/Conversion, conversion rate
• Sidebar filters: date range, campaign, ad group
• Trend chart for clicks and conversions over time
• Performance tables and charts by campaign and by ad group
• Period-over-period comparison with absolute and percentage deltas
• Download summarized results as CSV
• Bundled synthetic sample data for an instant demo

Data expectations

The dashboard expects at least a date column and one or more of: impressions, clicks, spend, conversions. Recognized date column names include date, day, Date. Campaign and ad group columns are optional but recommended. Any numeric column that is missing is treated as zero; derived metrics are never invented — they are computed from the columns actually present.

Technology stack

• Python 3.11
• Pandas (data processing)
• Plotly (interactive charts)
• Streamlit (web UI)
• openpyxl (XLSX upload support)

Folder structure

ppc-performance-dashboard/
├── app.py                  # Streamlit application
├── data_processor.py       # Normalization, KPI, filter, aggregation, comparison engine
├── tests/
│   └── test_data_processor.py
├── sample_data.csv         # Synthetic demo data (4 weeks, 4 campaigns)
├── requirements.txt
└── .gitignore

Installation

python -m venv venv
# Windows: venv\Scripts\activate  |  macOS/Linux: source venv/bin/activate
pip install -r requirements.txt

Usage

streamlit run app.py

Choose Use bundled sample data for an instant demo, or upload your own campaign export.

Testing

pytest -q

Tests cover column normalization, KPI math, campaign filters, date-range filters, aggregation, and period comparison.

Limitations and known issues

• The comparison assumes calendar-aligned equal-length periods
• Real exports vary by ad platform; unknown columns are ignored (no guessing)
• No multi-touch attribution or incrementality modeling — the dashboard reports platform-reported metrics as-is

Future improvements

• Budget pacing / burn-rate alerts
• Multi-platform data merge (Google Ads + Meta + Bing)
• Forecasts using simple time-series models
• Scheduled email delivery of the report

License

MIT (add a LICENSE file before commercial use if required by your use case).

---

Developed by Muhammad Hamid

Digital Marketing | SEO | Performance Marketing | E-commerce | Web Development

• Portfolio: https://abouthamid.web.app/
• LinkedIn: https://www.linkedin.com/in/muhammad-hamid-digital-marketing/
• WhatsApp: https://wa.me/923167417135

Have a project, job opportunity, or collaboration in mind? Contact Muhammad Hamid on WhatsApp.