ROAS and Media Budget Forecasting for an Agency
Advertising & Media — Planners had no way to predict how shifting budgets across Google, Meta and marketplaces would change revenue for their clients.…
Client: Performance Marketing Firm. Sector: Advertising & Media.
The challenge: Planners had no way to predict how shifting budgets across Google, Meta and marketplaces would change revenue for their clients.
What we built: A ROAS and marketing-mix forecasting model trained on two years of campaign history, plus a budget-scenario tool that recommends how to split spend across channels.
Tech used: Python, Bayesian MMM, BigQuery, Airflow, Looker Studio.
- Return on Ad Spend: +17%
- Planning Time: Days → hours
- Clients Added: 40+
- Forecast Error: <10%