Learning From Sales Data — Data Analysis Project

This project demonstrates how a realistic micro‑business sales dataset can be analysed to reveal clear, practical insights. Showing the next steps after cleaning: turning structured data into meaningful understanding that supports decision‑making.

Scenario

A small independent homeware shop wanted to understand what their sales data was really showing — which products performed best, how customers behaved, and where they could improve stock decisions. Their data lived in spreadsheets and monthly notes, but nothing was joined up or easy to interpret.

I cleaned the data lightly, structured it, and built a clear Power BI dashboard to reveal the patterns behind everyday sales.

Before → After

Raw dataset:

The dataset included:

  • inconsistent product names

  • mixed date formats

  • varied category labels

  • untidy notes

  • small structural issues

It was usable, but not analysis‑ready.

Dashboard Overview:

The dashboard includes:

  • YTD revenue, volume, averages

  • Category performance

  • Top products

  • Payment method breakdown

  • Customer type comparison

  • Notes themes

  • Clear, simple visuals designed for small‑team decision‑making

What I Did

  • Cleaned and organised the raw data

  • Structured the dataset for analysis

  • Built a Power BI dashboard

  • Created clear KPIs (revenue, volume, averages)

  • Highlighted top products by revenue and volume

  • Analysed customer behaviour (walk‑in vs returning)

  • Summarised payment method patterns

  • Extracted themes from customer notes

  • Designed clean visuals for quick decision‑making

Key Insights

  • Soft Furnishings generated the highest revenue, driven by consistent sales of wool throws.

  • Home Fragrance had the highest volume, especially scented candles and diffusers.

  • Returning customers spent significantly more per transaction than walk‑ins.

  • Card payments dominated, suggesting customers prefer quick, frictionless checkout.

  • Notes revealed recurring themes: gift wrap requests, delivery questions, accessibility needs, and reassurance about products.

Why This Matters

Once analysed, the dataset becomes genuinely useful. Oak & Ivy can now:

  • plan stock more accurately

  • understand customer behaviour

  • identify high‑value products

  • improve customer experience

  • make clearer decisions based on real patterns

Clear analysis supports clearer thinking.

Typical outcomes

  • Dashboards that show what matters at a glance

  • Monthly reporting that takes minutes, not hours

  • Better understanding of customer behaviour

  • Confident stock planning

  • Decisions based on real data, not guesswork

GitHub Link

View the full project on GitHub: https://github.com/Dawn-ThinkCave/sales_data_analysis