How do you take a mountain of raw data and turn it into something your boss actually wants to look at? That is the core problem this chapter solves: choosing the right chart, graph, or table so that your report tells a clear story without making your audience work to understand it. For the PL-300 exam, this matters because Microsoft tests your ability to pick the correct visual for a specific business question, not just your ability to click buttons.
Jump to a section
A simple way to picture Creating Visuals and Designing Reports
Because you first need to gather all your ingredients and tools before you can plate a beautiful meal, laying out a clear report structure naturally leads to selecting the right visuals for your audience.
Imagine you are renovating a kitchen. You don't start by randomly throwing in a fridge, a stove, and a sink. First, you figure out how you will use the kitchen. Will you be baking elaborate cakes (detailed trend analysis) or just reheating takeaway (simple KPI dashboards)? For a bakery kitchen, you need a powerful, gas-powered oven (a line chart showing sales over time). For a microwave-only kitchen, you just need a simple countertop model (a single number card showing total revenue). You also need to consider who is using the kitchen. If you are cooking for a professional chef (an executive), you might install a commercial-grade range (a complex scatter plot with outliers highlighted). If you are cooking for a toddler (a junior analyst), you would use child-safe locks and bright, simple knobs (a clean, colour-coded bar chart with no confusing filters). Finally, the layout matters. You wouldn't put the fridge next to the oven (mixing a map with a table that has no geographic context) because that creates a confusing workflow. Instead, you organise the work triangle: sink, fridge, and stove in a logical flow. In Power BI, this means placing a slicer for date range next to the relevant time-series chart, not at the bottom of a disconnected page.
When you create a report in Power BI, you are not just making it look pretty. You are building a visual translation layer between raw numbers and human understanding. The core concept here is 'mapping data fields to visual properties'. Every visual has properties like the X-axis (horizontal axis), Y-axis (vertical axis), legend (colour-coded categories), and values (the numbers being measured). You assign your data fields to these properties. For example, a bar chart has a category field (like Product Name) mapped to the axis, and a value field (like Total Sales) mapped to the height of the bars.
Why do we have different visuals? Because the human brain is terrible at reading raw tables of numbers but excellent at spotting patterns in shapes, colours, and positions. A line chart is perfect for showing trends over time because we naturally read left to right and see the ups and downs. A pie chart shows parts of a whole (proportions) but only works well with a few categories (like five or fewer). A stacked bar chart lets you compare totals while also seeing how each sub-category contributes.
Power BI offers these core visual types, each with a specific job:
Bar and column charts: Use these when you want to compare values across categories (e.g., sales by region). Column charts work best for fewer categories; bar charts (horizontal) are better when you have long category names.
Line charts: Use these when you want to show a trend over a continuous period (e.g., monthly revenue). They connect data points to show the direction of change.
Pie and donut charts: Use these only to show how a whole is divided into parts, and limit to 2-5 slices. Avoid them for comparing categories (a bar chart is better for that).
Maps (including filled maps): Use these when you have geographic data (country, state, city). A filled map shades regions based on a value, perfect for showing density (e.g., population per state).
Scatter plots: Use these to show the relationship between two numerical measures (e.g., advertising spend vs. website traffic). They can also include a third dimension via bubble size.
Tables and matrices: Use these when users need exact numbers. A table is simple rows and columns. A matrix adds row and column groupings, like a pivot table in Excel.
Cards and multi-row cards: Use a single card to display one key metric (e.g., total revenue). Use a multi-row card to show a few related numbers without a full table.
Slicers: These are interactive filters that let report viewers select which data they want to see (e.g., a month slicer). They are not visuals themselves but essential controls.
Gauges and KPIs: Use a gauge to show progress towards a target (e.g., sales vs. goal). A KPI visual adds a trend arrow and status colour.
Ribbon charts, waterfall charts, funnel charts: These are more specialised. A waterfall is great for showing how an initial value gets increased or decreased by a series of factors (e.g., profit calculation). A funnel shows attrition stages (e.g., sales pipeline).
The key to selecting the right visual is understanding your data's structure and your audience's question. If the question is 'What is the trend over time?', you pick a line chart. If the question is 'Which product sells the most?', you pick a bar chart. If the question is 'Which region has the highest profit?', you pick a map or a bar chart sorted by profit. Never use a pie chart to compare values — that traps your audience into comparing angles and areas, which humans are very bad at.
Finally, remember that Power BI applies default aggregation to numeric fields (usually SUM). If your data has granular rows (like individual transactions), you may need to change aggregation to COUNT, AVERAGE, or DISTINCT COUNT depending on what you are trying to show. For example, to show the number of customers, use DISTINCT COUNT of CustomerID, not SUM.
Identify the Business Question
Before you drag any fields, clarify what the report page should answer. Is it 'Show me monthly sales trend' or 'Compare sales by region'? This determines visual type.
Map Data Fields to Visual Roles
Open the Visualisations pane and select the chart type. Drag the relevant columns into the Axis (categories), Values (numbers), Legend (colours), and Tooltips (extra info). Power BI will create a default visual.
Apply Formatting for Clarity
Use the Format pane to add data labels, adjust colours (use company brand colours), change axis titles, and set number formatting (e.g., currency symbol, decimal places). Do not leave defaults that look generic.
Add Interactivity with Slicers and Filters
Drag a date or category field onto the canvas and select the Slicer visual type. Configure it as a drop-down or list. This lets users filter the entire page without editing the underlying data.
Enable Drill-Down or Drill-Through
If your visual uses a hierarchy (e.g., Year > Quarter), click the drill-down icon (double arrow) to allow users to click to see deeper levels. For drill-through, create a separate page and set field parameters in the Drill-through pane.
Test Cross-Filtering and Sorting
Click on a bar in the visual and verify that other visuals on the page update to show only the filtered subset. Also, check that the sort order (ascending/descending) matches the audience's expectation (e.g., highest first).
Publish and Monitor Performance
Upload the report to Power BI Service. Check the Performance Analyzer (in Power BI Desktop) before publishing to ensure no visual takes too long to load. Large tables or complex measures may need optimisation via summarised data.
Let us walk through a real day for Maria, a junior data analyst at a mid-sized retail company. She has been asked to build a monthly sales performance report for the head of marketing, who is very busy and hates reading long tables.
Maria starts by connecting to the company's sales database which contains columns: OrderDate, ProductCategory, SalesAmount, CustomerCity, and SalesRepName. She knows the marketing head needs answers to three questions: (1) How did total sales perform this month compared to last month? (2) Which product categories are growing? (3) Which cities have the highest sales?
For question one, Maria decides against a table because the executive needs a quick 'at a glance' answer. She inserts a card visual to show total sales for the current month ('$1.2M') and a mult-row card that also shows last month's value. She adds a gauge to show progress against the monthly target ('88% of $1.5M target'.
For question two, she needs to show trend and comparison. She drags SalesAmount into a line chart and OrderDate to the axis, then sets the date level to 'Month'. She sees a steady upward slope. To compare categories, she adds ProductCategory to the legend. Now each category has its own coloured line. She notices 'Electronics' is flat but 'Home Goods' is rising sharply. That is actionable insight.
For question three, she considers a map. She drags CustomerCity to the location field and SalesAmount to the colour saturation field. A filled map of the UK appears, with London shaded very dark (high sales) and other cities lighter. She then adds a slicer for SalesRepName at the top of the report so the marketing head can filter by individual sales rep if needed.
Maria is careful about clutter. She does not use a pie chart for 15 product categories because it would be unreadable. She does not use a scatter plot because she has no continuous numerical X and Y variables. She keeps the layout simple: the card and gauge at the top (summary), the line chart in the middle (trend), and the map on the right (geography). She adds a slicer on the left. She also formats the map colour to use a green gradient (green = good) instead of default red, because her audience associates green with positive performance.
Finally, she publishes the report to the Power BI service and sets it to refresh daily. She also pins the card and gauge to a dashboard so the marketing head sees them first thing in the morning without opening the full report.
The PL-300 exam tests the objective 'Select and create appropriate visualizations for reports' heavily. Expect 5-10 questions that present a business scenario and ask you to choose which visual best answers the question. Microsoft loves to test 'why' you pick one visual over another.
Common question patterns:
Scenario-based selection: The question will describe a business user's need. For example: 'A sales manager wants to see the trend of monthly sales over the past year and identify the month with the highest sales.' The correct answer is a line chart with markers (clearly shows trend and peak). A column chart is wrong because it shows discrete values, not a continuous trend line.
Data type traps: They will ask you to match a visual to a data type. For example: 'You have a column with continuous numerical values (e.g., temperature). Which visual is appropriate?' Answer: a histogram (if showing distribution) or a line chart (if over time). A bar chart requires categorical data, so it would be wrong for continuous numbers.
Limitations of visuals: They love to test when not to use a visual. For example: 'You have 20 product categories. Which visual should you avoid?' Answer: pie chart (because humans cannot compare 20 slice areas accurately). Another trap: 'You have geographic data but no longitude/latitude.' Solution: use a filled map with country/state names, not a map that requires coordinates.
Interaction and hierarchy: Questions about drill-down and drill-through. They will ask: 'Which visual supports hierarchical drill-down (e.g., year > quarter > month)?' Answer: any chart that supports drill-down (matrix, bar chart with date hierarchy). A card does not.
Filtering and slicers: They might ask: 'Which visual allows users to filter other visuals on the same page?' Answer: a slicer. They might also ask about 'sync slicers' across pages.
Analytics pane: They ask about adding trend lines, constant lines, or forecasting. The analytics pane is available for line and scatter charts, not for bar charts (in most cases).
Performance considerations: For large datasets, avoid using too many custom visuals or complex visuals like scatter plots with many data points. The exam might ask which visual is most efficient.
Accessibility: They test alternative text for visuals and ensuring colour choices are colour-blind friendly (use built-in colour-blind palettes).
Key definitions to remember:
Axis: X-axis (horizontal, usually categories or time), Y-axis (vertical, usually values).
Legend: Explains what each colour or shape represents.
Drill-down: Clicking a visual to see more detailed data (e.g., from year to quarter).
Drill-through: A filter that sends a context (e.g., a selected product) to another report page.
Cross-filtering: Clicking a bar in one chart filters the other charts on the page.
Hierarchy: A parent-child relationship in data (e.g., Year > Month > Day).
Always match the visual to the specific business question: trend = line chart, comparison = bar chart, proportion = pie chart (2-5 slices), geographic density = filled map.
Limits: never use a pie chart for more than five categories, and never use it to compare values across categories (use a bar chart instead).
Use card visuals to display a single, crucial KPI like total revenue or current headcount; avoid cluttering the page with multiple cards for minor metrics.
Slicers allow report viewers to interactively filter the entire page; place them at the top or left edge for consistency.
Cross-filtering is enabled by default: clicking a bar in one chart will filter other charts on the same page; you can disable this in the Format pane if needed.
For accessibility, use high-contrast colour palettes and always provide alternative text descriptions for each visual.
Drill-down lets users explore hierarchies (e.g., year to quarter) directly in the visual; set up a date hierarchy in your data model to enable this.
The Analytics pane (trend lines, constant lines, forecasting) is only available for line, scatter, and area charts, not for bar or column charts.
When comparing parts of a whole over time, use a 100% stacked bar chart (each bar sums to 100%) rather than a pie chart for each time period.
Custom visuals from AppSource are powerful but can slow down report performance; only use them when a built-in visual cannot meet the requirement.
These come up on the exam all the time. Here's how to tell them apart.
Line Chart
Best for showing continuous trends over time (e.g., monthly revenue).
Connects data points with a line to highlight direction and rate of change.
Supports multiple series (lines) for comparison, but can become cluttered with more than 4 lines.
Column Chart
Best for comparing discrete categories (e.g., sales by product).
Uses vertical bars; the height of each bar represents the value.
Can be clustered to show sub-categories side-by-side, but does not show a smooth trend.
Pie Chart
Shows parts of a whole as slices of a circle.
Difficult to compare slice sizes accurately, especially with more than 5 slices.
Better when you want to emphasise the largest share visually.
100% Stacked Bar Chart
Shows parts of a whole as stacked sections within a horizontal bar.
Easier to compare across multiple categories (each bar is a separate category).
Better when you have many categories (5+) or want to compare composition over time.
Slicer
Visible on the report page; users can see and interact with it directly.
Can be formatted as a dropdown, list, or tile.
One slicer can filter all visuals on the page (by default).
Filter Pane (page-level filter)
Hidden from the report canvas; only visible in the Filters pane (editors).
Cannot be formatted for appearance; it is purely functional.
Does not visually show which options are currently selected; user must open the pane.
Table Visual
Simple rows and columns with no grouping.
Best for showing detail-level transactional data.
Does not support subtotals by default.
Matrix Visual
Supports row and column groupings (like a pivot table).
Automatically calculates subtotals and totals for each group.
Allows drill-down within hierarchies (e.g., expand a year to see quarters).
Mistake
Pie charts are always bad and should never be used.
Correct
Pie charts are acceptable when showing 2-5 categories representing parts of a whole and the audience needs to see proportional differences, not precise comparisons.
Beginners read blanket advice that 'pie charts are evil' without understanding their specific use case, so they avoid them even when appropriate.
Mistake
You should always use the default visual that Power BI suggests when you drop fields onto the canvas.
Correct
Power BI's auto-suggested visual is often a table or a generic column chart, which may not be the most effective for the specific question you are answering.
Beginners trust the tool's default suggestions because they seem 'smart', but the tool does not know the business context or audience needs.
Mistake
A map is the best visual for any data that contains a city or country name.
Correct
Maps are only useful if the audience needs to see geographic patterns (e.g., density or regional variation). If you just need to compare sales numbers across cities, a sorted bar chart is more efficient.
Maps look impressive and 'data-sciencey', so beginners gravitate towards them even when the data question does not involve geography.
Mistake
More visuals on a page make the report more informative.
Correct
Adding too many visuals creates clutter and makes it harder for the audience to find the key insight. The best reports focus on one or two key messages per page.
Beginners fear that a single chart is 'not enough' and try to cram every possible angle onto one page, not realising that simplicity drives understanding.
Mistake
All numeric fields in a visual must use SUM aggregation.
Correct
Depending on the question, you may need COUNT, AVERAGE, DISTINCT COUNT, MIN, or MAX. For example, average order value uses AVERAGE, not SUM.
Beginners do not realise that Power BI applies a default aggregation (SUM) and assume that is always correct, leading to misleading totals.
Reveal each answer, then mark whether you got it right. Score 60%+ to unlock the next chapter.
A table is a simple grid of rows and columns with no grouping. A matrix adds row and column groupings (like a pivot table) and can show subtotals, making it better for hierarchical data like year-month-product.
Gaps appear when there are missing date values in your data (e.g., no sales on a weekend). You can connect the gaps by turning on 'Show items with no data' in the field settings, or remove the gaps by switching to a continuous axis.
You can, but a line chart is better because the continuous line helps the eye see the direction and speed of change more easily than comparing discrete bar heights.
Drag multiple measures (e.g., Sales, Costs) to the Values field well. The chart will add a second axis or combine them into a clustered bar chart. For better clarity, use a combination chart (column + line) from the visual gallery.
It lets you add a trend line (linear or moving average), a constant line (e.g., a target), or a forecast line. These are only available for line, scatter, and area charts.
First, create a date hierarchy in the data model (Year > Quarter > Month). Then drag the hierarchy to the Axis of a bar or column chart. The drill-down buttons (up/down arrows) will appear in the top-left corner of the visual.
Cross-filtering means clicking on a visual filters other visuals on the same page. To turn it off for a specific visual, go to Format > Edit Interactions and select the 'no interaction' icon (slashed circle) for that visual.
You've finished Creating Visuals and Designing Reports. Continue through the PL-300 study guide to build a complete picture of the exam.
Done with this chapter?