Café Revenue Analysis

Executive summary

A Python analysis of 232 anonymised café transactions from April 2026. The work turns handwritten records into an operational view of revenue mix, customer credit, service days, payment methods, and pricing discipline.

Business context

The Harare business traded through walk-ins, account customers, and room hire but had no digital transaction system. The owner could not see which stream carried the business, where credit exposure sat, or whether service days were worth the effort.

Data and preparation

Notebook records were digitised into Excel with AI-assisted transcription and then checked row by row. Customer identifiers and financial figures were anonymised before the pandas analysis. One month is enough to diagnose control gaps, but not enough to claim seasonality or margin trends.

Method

Transactions were grouped by revenue stream, account customer, day, payment method, item, and room-hire status. pandas, matplotlib, seaborn, and notebook-based checks were used to test customer concentration, daily lift, and pricing consistency.

Results, limits, and next step

Walk-in cash trade contributes roughly two-thirds of revenue. Card payments account for under one percent of transactions, one account customer paid less than a quarter of charges, and service days lift daily revenue by about fifteen percent. The clearest recommendation is a point-of-sale system to lock pricing, track credit, and make card payment normal. Cost data is the main missing piece for a margin view.