1. Dataset read | Variable | Type | Role in analysis | Key relationships worth visualising | Priority for goal | | --- | --- | --- | --- | --- | | order_id | Discrete identifier | Transaction grain | Order frequency, duplicate checks | Low | | order_date | Temporal | Time axis | Trends, seasonality, growth rate, anomalies | Highest | | product_category | Nominal categorical | Revenue driver segmentation | Revenue contribution, mix shifts, category trends | Highest | | region | Nominal categorical | Geographic segmentation | Regional comparisons, market concentration, trend differences | Highest | | units | Continuous (count-like) | Volume metric | Volume vs revenue, outliers, demand patterns | Medium | | unit_price | Continuous | Pricing metric | Revenue composition, pricing distribution, premium vs low-cost categories | Medium | | revenue | Continuous | Core KPI | Trend, distribution, concentration, contribution analysis | Highest | | customer_id | Discrete identifier | Customer behaviour | Repeat purchase patterns, customer concentration | Medium | Relationships that matter most Revenue over time → identifies growth, seasonality, and shocks. Revenue by category → identifies commercial drivers. Revenue by region → identifies geographic performance differences. Category × region interaction → reveals where categories succeed or fail. Units vs revenue / price vs revenue → separates pricing effects from demand effects. Revenue concentration → detects overreliance on a few categories or regions. 2. Chart recommendations | Relationship | Recommended chart | Why (perceptual + analytical) | Alternative | What to watch out for | | --- | --- | --- | --- | --- | | Revenue trend over time | Line chart | Position along a common scale is the most accurate perceptual encoding; ideal for trend and seasonality detection | Area chart | Avoid dual axes unless units are standardised | | Revenue by product category | Sorted horizontal bar chart | Length comparisons are highly accurate; executives scan ranked categories quickly | Treemap | Too many categories reduce readability | | Revenue by region over time | Small-multiple line charts | Preserves trend shape while reducing overlap clutter; enables regional comparison | Multi-line chart | Shared axes required for fair comparison | | Category contribution to total revenue | 100% stacked bar chart | Shows changing composition over time; easy part-to-whole reading | Stacked area chart | Hard to compare middle segments precisely | | Revenue distribution by category | Box plot | Displays spread, skew, and outliers compactly | Violin plot | Outliers may dominate scaling | | Units vs revenue | Scatter plot with regression line | Reveals relationships and clusters; useful for separating volume-driven vs price-driven growth | Hexbin plot | Overplotting with 50k points | | Region × category revenue intensity | Heatmap | Color intensity supports matrix scanning efficiently | Highlight table | Sequential palette must be perceptually uniform | | Monthly revenue change | Waterfall chart | Shows additive contributions to growth/decline clearly | Diverging bar chart | Ensure clear baseline and ordering | | Top customers concentration | Pareto chart | Combines ranking and cumulative share insight | Lorenz curve | Executive audiences may misread cumulative axis | Recommended executive-core chart set For a C-suite audience, prioritise: Revenue trend line chart Revenue by category bar chart Regional small multiples Category-region heatmap KPI cards with sparklines These maximise interpretability and reduce cognitive load. 3. Colour palette Categorical palette (colour-blind safe, WCAG 2.1 AA-aware) | Purpose | HEX | | --- | --- | | Blue | #0072B2 | | Orange | #E69F00 | | Green | #009E73 | | Vermillion | #D55E00 | | Purple | #CC79A7 | | Sky blue | #56B4E9 | | Yellow | #F0E442 | | Dark gray | #4D4D4D | Rationale: Based on proven colour-blind-safe palettes (similar to Okabe-Ito); high hue separation improves categorical discrimination and executive readability. Sequential palette (for heatmaps) | Step | HEX | | --- | --- | | 1 | #F7FBFF | | 2 | #C6DBEF | | 3 | #6BAED6 | | 4 | #3182BD | | 5 | #08519C | | 6 | #08306B | Rationale: Monotonic luminance progression improves magnitude perception and preserves readability in grayscale printing. Grayscale fallback | Tone | HEX | | --- | --- | | Light | #D9D9D9 | | Medium-light | #BDBDBD | | Medium | #969696 | | Medium-dark | #636363 | | Dark | #252525 | Use for: printed reports, accessibility fallback, monochrome exports. 4. Dashboard layout Executive dashboard (recommended) Design principles Z-pattern reading: KPIs first, trends second, diagnostics last. Minimal interaction burden. Large typography for key metrics. One primary message per chart. ASCII wireframe +-----------------------------------------------------------+ | KPI Revenue | KPI Growth % | KPI Avg Order | KPI Regions | | Sparkline | Sparkline | Sparkline | Sparkline | +-----------------------------------------------------------+ +-----------------------------+-----------------------------+ | Revenue Trend Over Time | Revenue by Category | | (line chart) | (sorted horizontal bars) | +-----------------------------+-----------------------------+ +-----------------------------+-----------------------------+ | Regional Trends | Region x Category Heatmap | | (small multiples) | | +-----------------------------+-----------------------------+ +-----------------------------------------------------------+ | Revenue Drivers Diagnostic | | Scatter: Units vs Revenue | +-----------------------------------------------------------+ Visual hierarchy KPIs → immediate business status Trend chart → macro performance Category and region → revenue drivers Diagnostic charts → operational interpretation Interaction recommendations Filters: date range, region, category Hover tooltips with exact values Default to monthly aggregation Avoid more than 6–8 visible categories at once 5. Example to build it Example: Revenue trend line chart in Python (matplotlib + seaborn) Python Execută # Import libraries import pandas as pd import matplotlib.pyplot as plt import seaborn as sns # Sample dataset (10 rows) data = { "order_date": [ "2025-01-01", "2025-01-15", "2025-02-01", "2025-02-15", "2025-03-01", "2025-03-15", "2025-04-01", "2025-04-15", "2025-05-01", "2025-05-15" ], "region": [ "North", "South", "North", "West", "East", "South", "West", "North", "East", "South" ], "revenue": [ 1200, 1500, 1700, 1600, 1800, 2100, 2200, 2400, 2300, 2600 ] } # Create DataFrame df = pd.DataFrame(data) # Convert dates to datetime df["order_date"] = pd.to_datetime(df["order_date"]) # Aggregate monthly revenue monthly_revenue = ( df.groupby(pd.Grouper(key="order_date", freq="M"))["revenue"] .sum() .reset_index() ) # Set visual theme sns.set_theme(style="whitegrid") # Create figure fig, ax = plt.subplots(figsize=(10, 5)) # Plot revenue trend sns.lineplot( data=monthly_revenue, x="order_date", y="revenue", marker="o", linewidth=2, ax=ax ) # Add chart title and labels ax.set_title("Monthly Revenue Trend", fontsize=16, weight="bold") ax.set_xlabel("Month") ax.set_ylabel("Revenue") # Improve date formatting fig.autofmt_xdate() # Add annotations as lightweight tooltips substitute for x, y in zip(monthly_revenue["order_date"], monthly_revenue["revenue"]): ax.annotate( f"${y:,.0f}", (x, y), textcoords="offset points", xytext=(0, 8), ha="center", fontsize=9 ) # Reduce unnecessary visual clutter sns.despine() # Display chart plt.tight_layout() plt.show() Additional implementation guidance Heatmap Use pivot_table(index="region", columns="product_category", values="revenue", aggfunc="sum") Plot with sns.heatmap(..., cmap=sequential_palette, annot=True) Small multiples Use sns.relplot(kind="line", col="region", col_wrap=2) Keep all y-axis limits identical Executive formatting Use: font sizes ≥ 12 pt limited gridlines direct labels where possible no 3D effects no rainbow palettes 6. How to read it For executives What the charts show Whether revenue growth is accelerating or slowing, and which categories/regions drive it. If you see X, do Y One category dominates growth → protect supply chain and margin exposure. A region lags while others grow → investigate pricing, logistics, or local competition. Revenue grows but units stay flat → check if growth is price-driven rather than demand-driven. Red flags Sharp seasonal drops Revenue concentrated in one region/category High volatility month-to-month Growth driven only by price increases For analysts Metric definitions Revenue = units × unit_price Monthly revenue should be aggregated at consistent calendar intervals. Cautions Correlation between units and revenue does not imply demand causation; pricing changes may dominate. Regional differences may reflect customer mix, not operational quality. Follow-up questions Are returns/refunds excluded? Do category trends differ by season? Is growth driven by new or repeat customers? For data scientists Statistical and modelling considerations Aggregate to monthly granularity for executive dashboards to reduce noise. Use log scaling if revenue distribution is highly skewed. Winsorise or inspect extreme outliers before regression analysis. Data-quality checks Duplicate order_id Missing dates or categories Negative revenue values Inconsistent regional labels Next-step ideas Customer segmentation (RFM: Recency, Frequency, Monetary value) Revenue forecasting with time-series models Category uplift modelling Regional clustering analysis Price elasticity estimation