Sales Data Analysis: Ama's Fashion Store
The task
Ama owns two clothing shops and wanted to know what to buy more of for her next stock order. Her sales sheet had 500 rows, but it was messy and couldn't be trusted as it was.
How I cleaned the data
- Removed 10 duplicate sales (500 → 490 rows)
- Fixed 4 wrong prices, e.g. a GHS 95 shirt typed as 950
- Standardised spellings ("MoMo" → "Mobile Money", "Top" → "Tops")
- Put dates written three different ways into one format
- Flagged 9 sales with no payment status for follow-up
- Kept a cleaning log of every change, so nothing was hidden
What the data showed
Ankara Dress earns the most money
Revenue by product (GHS)
…but Linen Shirt sells the most items
Units sold by product
Sales grew every month
Revenue by month (GHS)
Most customers pay with Mobile Money
Share of revenue by payment method
- Stock more Ankara Dresses and Linen Shirts. One brings in the most money, the other sells the most items.
- Collect GHS 8,495 in unpaid or unconfirmed sales before the next order.
- Look into why Osu earns GHS 20,755 less than East Legon before shifting stock between shops.
- Keep Mobile Money fast and reliable at both shops, since it carries 62% of revenue.
Tools I used
Training project completed at Soronko Academy using a sample business dataset.