Revenue Forecasting From Financial Data
Budget: €18 – €36 EUR
I have several years of detailed financial records—monthly P&L statements, sales ledger exports, and a handful of operational KPIs—and I need them turned into a clear, defensible revenue forecast that will drive my upcoming budget.
The core of the job is budgeting and forecasting, with revenue forecasting as the specific focus. I’ll supply the raw data in CSV and Excel. From there, I need:
• A cleaned, well-structured dataset (Excel or Python/Pandas is fine)
• A forecasting model that explains its own logic—ARIMA, Prophet, or an equivalent method is acceptable as long as assumptions are transparent
• A concise report or dashboard (Power BI, Tableau, or Google Data Studio) that visualises projected revenue, highlights variance against historical trends, and allows simple scenario tweaks
Acceptance criteria
• Forecast spans at least the next 12 months with monthly granularity
• Mean Absolute Percentage Error on back-tests under 10 %
• All formulas or code fully commented so I can maintain or extend the model later
Everything should be delivered in a single shared folder—dataset, model files, and the final visual report—so I can drop it straight into my planning workflow.
The core of the job is budgeting and forecasting, with revenue forecasting as the specific focus. I’ll supply the raw data in CSV and Excel. From there, I need:
• A cleaned, well-structured dataset (Excel or Python/Pandas is fine)
• A forecasting model that explains its own logic—ARIMA, Prophet, or an equivalent method is acceptable as long as assumptions are transparent
• A concise report or dashboard (Power BI, Tableau, or Google Data Studio) that visualises projected revenue, highlights variance against historical trends, and allows simple scenario tweaks
Acceptance criteria
• Forecast spans at least the next 12 months with monthly granularity
• Mean Absolute Percentage Error on back-tests under 10 %
• All formulas or code fully commented so I can maintain or extend the model later
Everything should be delivered in a single shared folder—dataset, model files, and the final visual report—so I can drop it straight into my planning workflow.