14. Visuals and Their Formatting
14.3 Choose the Right Chart
Quick answer: the short, self-contained version of this lesson is the guide How to Choose a Chart in Power BI.
Start with a problem
Zoya must present three questions to Rani: which city sells most, how sales moved from April to May (FY 2026-27), and what share each category has. She is about to put all three in pie charts.
Explain simply
Pick the chart from the question: comparing items, then bar or column; change over time, then line; part of a whole with few parts, then donut or 100% bar; two measures across many items, then scatter; one number, then card; number against a target, then KPI or gauge.
Show the data
| Question | Data needed |
|---|---|
| Which city sells most? | Total Sales by City |
| How did sales move Apr, then May? | Total Sales by Month |
| What share does each category have? | Total Sales by Category |
About this data: This table uses columns: Question, Data needed. This table uses columns: Question, Visual, Answer. This table uses columns: Question type, Best visuals, Quick-commerce example, Avoid. This table uses columns: Wrong result, Reason, Correction. Values shown are the lesson examples already on this page (not a new dataset).
Fictional sample data. Practice file: visuals_orders_apr_may_2026.csv (source: src/practice/).
Demonstrate
- Write each question in one sentence on paper.
- Underline the word that tells the shape: most (compare), move (trend), share (part of whole).
- Insert a Clustered bar chart for the city question.
- Insert a Line chart with
Order Date(Month level) for the trend question. - Insert a Donut chart with
Categoryfor the share question.
Show the result
Your result should look like this (fictional sample):
| Question | Visual | Answer |
|---|---|---|
| Which city sells most? | Bar | Pune ₹850 |
| Apr, then May trend | Line | ₹900, then ₹1,200 (+33.3%) |
| Category share | Donut | Dairy 50.0%, Snacks 31.0%, Fruits 19.0% |
Explain what happened
Each visual uses the same Total Sales measure. Only the field on the axis changes, so the filter context changes: per city, per month, per category.
Explore further
- Use a table when people must read exact values.
- Avoid 3-D and more than five donut slices.
Detailed reference for this lesson:
Start with the question, then pick the visual.
| Question type | Best visuals | Quick-commerce example | Avoid |
|---|---|---|---|
| Compare categories | Clustered bar/column | Orders by City; Sales by Category | Pie with many slices, 3D effects |
| Trend over time | Line, area | Daily orders (Diwali/Ganeshotsav spikes); Sales vs Sales LY | Bars for 365 days |
| Composition (part-to-whole) | Stacked/100% stacked bar, donut (2–4 parts), treemap | Payment Mode mix per City; Platform share | Pie for > 5 categories |
| Change from start to end | Waterfall | Gross sales, then discounts, then delivery fees, then net | Line chart |
| Relationship between two measures | Scatter / bubble | Orders vs Avg Delivery Time per dark store | Two separate bar charts |
| Distribution | Column chart of bins (Module 12.6), box-plot custom visual | Delivery time in 5-min bins | Average only |
| Ranking change over time | Ribbon chart | Category rank by month | Many lines crossing |
| Process stages | Funnel | Placed › Packed › Out for delivery › Delivered | Pie |
| Single KPI | Card, Card (new), KPI, gauge | AOV; % Under 10 Mins vs target | A table with one cell |
| Geography | Map / Azure map / filled map (Module 15) | Orders by dark store location | A map when location is not the point |
| Details / exact numbers | Table, matrix | Order list; City × Month matrix | Charts with 50 labels |
| Root cause | Decomposition tree, key influencers | Why are Nagpur deliveries slow? | Guesswork |
He table phone madhe photo kadhun theva – kontahi chart nivadtana kaamala yeil.
Ravindra Bagale's Tip
A common mistake is choosing a chart because it looks attractive instead of because it fits the question. Use the table in this section: comparison, then bar, trend, then line, part-to-whole with few categories, then donut or 100% bar, relationship, then scatter. When unsure, a sorted bar chart is almost always readable. Try it once more, then move on.
Ravindra Bagale's Tip – मराठी
एक common चूक म्हणजे प्रश्नाला जुळतो म्हणून नाही तर आकर्षक दिसतो म्हणून chart निवडणं. या section मधला table वापरा: तुलना, then bar, trend, then line, कमी categories असलेलं part-to-whole, then donut किंवा 100% bar, संबंध, then scatter. खात्री नसेल तर sorted bar chart जवळजवळ नेहमीच वाचता येतो. पुन्हा एकदा करून बघा, मग पुढे जा.
Ravindra Bagale's Tip – हिंदी
एक common गलती है chart इसलिए चुनना कि वह आकर्षक दिखता है, इसलिए नहीं कि वह सवाल पर फ़िट बैठता है. इस section की table इस्तेमाल करो: तुलना, then bar, trend, then line, कम categories वाला part-to-whole, then donut या 100% bar, संबंध, then scatter. पक्का न हो तो sorted bar chart लगभग हमेशा पढ़ने लायक होता है. एक बार फिर करके देखो, फिर आगे बढ़ो.
Fix common mistakes
| Wrong result | Reason | Correction |
|---|---|---|
| A pie with 15 slices nobody can read | Part-of-whole chart for many items | Use a sorted bar chart |
| Line chart with City on the axis | Line implies continuity in time | Use a bar chart for categories |
| Trend shows Jan before Apr in an FY report | Calendar sort, not FY | Sort by an FY month number column (Apr = 1) |
Practise independently
Practice task
For the question Did Amazon Now or Blinkit earn more? choose a visual and give the answer.
Hints
- Two items compared, then bar or column.
- Sum Amount by Platform.
Solution
Clustered column chart: Blinkit ₹1,100, Amazon Now ₹1,000.
Think like a professional
Interviewers often ask how do you choose a chart? Answer with the question-first rule and one example from your project. It shows you design for decisions.
Revision card
Interview card