Lesson 3 of 16 · Level 1: Beginner

Power BI Visuals, Filters and Slicers

  • ⏱ 12 min read
  • ❓ 4 review questions
  • ✍️ Exercise with solution
In this lesson
  1. Choose the chart based on the question
  2. The most useful visuals
  3. Filters: four levels
  4. Interactions between visuals
  5. Formatting that communicates
  6. Case study: where does Summit sell the most?
  7. Common visualization mistakes
  8. Summary
What you'll learn
  • Pick the right visual for each business question.
  • Use cards, bar and line charts, matrices, maps and combo charts.
  • Filter at the visual, page and report level, and with slicers.
  • Format visuals so the message lands in seconds.
  • Analyze Summit's sales by region and channel with the right chart.

A good report isn't the one with the most charts; it's the one that answers the important questions at a glance. In this lesson you'll learn to choose and set up Power BI visuals using one simple test: what question does this chart answer?

39.7%of Summit's sales come from the wholesale channel: $204,586 out of $515,969
19.8%comes from Chicago, the top city, ahead of New York (17.0%)
4levels of filtering: visual, page, report and slicer

Choose the chart based on the question

QuestionVisualSummit Supply example
How much? (one key number)Card, multi-row card, KPISales this month, number of customers, gross margin %
Who sells the most? (compare categories)Horizontal bars, sortedSales by rep or by product
How is it trending?Line chart (or columns for a few periods)Monthly sales over the last 24 months
What's it made of?Stacked bars, waterfall; a donut only for 2–4 partsSales by category within each city
Actual vs. target?KPI, clustered columns, line and clustered column chartSales vs. target by month
Where?Map or filled mapSales by city or state
Is there a relationship?Scatter chartAverage discount vs. margin by customer
What are the details?Matrix or tableSales by product and month with totals
Avoid pie charts with many slices

People are bad at comparing angles. With more than 4 categories, a sorted bar chart is always clearer. And skip 3D effects: they distort proportions.

The most useful visuals

Cards and KPIs

A card shows one big number. Use 3 to 5 of them across the top of a report for the headline figures. The KPI visual shows a value, its target and its trend, colored green or red. Format numbers so they're readable (thousands or millions) without decimals nobody needs.

Bar and column charts

Sort them from highest to lowest (the visual's "More options" menu, the three dots, › Sort axis). With many categories, use horizontal bars so labels stay readable. Turn on Data labels and remove the axis line if the labels already show the values.

Line charts

For trends over time. Add a trend line or an average line from the visual's Analytics pane. To compare two years, put Year in the Legend and Month on the X-axis.

Matrix

The matrix is Power BI's pivot table: rows, columns and values, with subtotals and expandable levels (Category → Product). It's ideal for people who need the numbers. Add conditional formatting (background color scales, data bars or icons) so the important values stand out.

Combo chart

The line and clustered column chart puts columns and a line on one visual, with a secondary axis: for example, sales as columns and gross margin % as a line.

1. You want to show which of 12 sales reps sold the most this quarter. Which visual do you use?

  • A pie chart with 12 slices.
  • A line chart.
  • A horizontal bar chart sorted from highest to lowest.

Comparing categories is easiest with sorted bars; 12 pie slices are impossible to compare.

Filters: four levels

LevelAffectsExample
Filters on this visualOnly that chartShow only the top 10 products (Top N filter)
Filters on this pageEvery visual on that pageA "Chicago" page filtered to City = Chicago
Filters on all pagesThe whole reportExclude canceled orders
SlicerThe page (and other pages if you sync it)On-canvas buttons so users pick a year, city or category

Filters in the Filters pane are usually set by the report author; slicers are for users to explore. You can hide or lock filters in the pane so users can't change them.

Slicer styles

List or dropdown

For cities, categories or reps.

Tip: a dropdown saves space when there are more than 6 options, like Summit's 11 cities.

Between

A slider for date or amount ranges.

Tip: perfect for letting users choose a start and end date.

Relative date

"Last 3 months" or "this year", which updates itself.

Tip: nobody has to remember to move the dates every month.

Tiles or buttons

For a few options, more visual.

Tip: great for Summit's 3 channels or 4 regions.

Recent versions of Power BI Desktop include newer slicers (button, list and text slicers) and a card visual that shows several values at once. Names can vary by version, but the logic is the same: the user picks, and the page filters.

With View › Sync slicers, one slicer can filter several pages.

2. You want no page of the report to include orders with the status "Canceled". What do you use?

  • A slicer on every page.
  • A report-level filter (Filters on all pages) that excludes "Canceled".
  • A filter on a single chart.

A report-level filter applies to every page, and users don't have to remember to apply it.

Interactions between visuals

By default, clicking an element in one chart filters or highlights the others. You can change that: select a chart, go to Format › Edit interactions, and choose for every other visual whether it filters, highlights or ignores the click. For example, a card with the annual total shouldn't change when someone clicks a month.

Formatting that communicates

  • Titles that state the takeaway when the report is for a presentation: "Chicago and New York bring in 37% of sales" instead of "Sales by city".
  • Less ink: remove gridlines, borders and backgrounds that add nothing.
  • Color with intent: one neutral color for everything and one accent color for what you want to highlight. Red and green only for bad and good.
  • Readable numbers: display units in thousands (K) or millions (M), thousands separators, no decimals that add nothing.
  • Consistency: the same color for the same category in every chart.

Case study: where does Summit sell the most?

The sales director asks which region sells the most. Show a pie chart and she'll see four nearly identical slices and can't decide anything. With sorted horizontal bars, the answer jumps out:

Sales by region, 2025–2026 (% of total)

Midwest25.9%
West25.6%
South25.2%
Northeast23.3%

The Midwest ($133,438), West ($131,912) and South ($130,217) are nearly tied; the Northeast ($120,402) trails the Midwest by about $13,000. The accent color on the Northeast draws the eye to what matters, and the chart title can say it outright: "The Northeast sells about 10% less than the Midwest." That's the power of choosing the right visual and color.

For the channel mix, on the other hand, a 100% stacked bar works well because there are only three parts:

Summit's sales by channel, 2025–2026: wholesale $204,586, convenience stores $186,037 and independent grocers $125,346.

Conditional formatting, step by step

In a matrix, click the drop-down arrow on the value field (or go to Format › Cell elements) and choose what to highlight:

  • Background color with a color scale: from white to blue by value. Great for spotting which product and month sell most.
  • Data bars: a bar inside each cell, combining the exact number with a visual comparison.
  • Icons: arrows or traffic lights with rules ("if the change is below 0, show a red down arrow").
  • Rules: colors for ranges you define, such as red when target attainment is below 90%.

Use one technique per column. A matrix with color scales, data bars and icons all at once turns into an unreadable rainbow.

3. You click a month in the line chart and the "Sales this year" card changes, but you wanted it to always show the annual total. What do you do?

  • Use Edit interactions so the line chart doesn't affect that card.
  • Delete the card.
  • Duplicate the page.

Edit interactions controls which visuals filter which.

4. In a column chart of sales by region, the axis starts at $115,000 and the Midwest column looks more than three times as tall as the Northeast. What's going on?

  • The Midwest sells three times as much as the Northeast.
  • The axis doesn't start at zero, which exaggerates the gap: bar and column axes should start at zero.
  • The Northeast data has an error.

Bar length is read against its baseline. Starting at $115,000 makes the Midwest ($133,438) and the Northeast ($120,402) look wildly different, when the real gap is about 10%.

Common visualization mistakes

A pie chart with 12 products, or with nearly equal slices.

Sorted horizontal bars; save donuts for 2 to 4 clearly different parts.

Truncating a bar chart's axis "so the difference shows".

Start bars at zero; if the gap is small, say so in the title.

Ten different colors for ten categories with no meaning.

One neutral color for everything and one accent for the point you're making.

Default titles like "Sum of Quantity by RepID".

Business-language titles: "Units sold by sales rep".

Filling the page with slicers "in case users need them".

Keep 2 or 3 key slicers on the canvas and the rest in the Filters pane.

Practice exercise: a page that answers questions

Using the data from the previous lesson, build one page that answers these five questions, each with the most suitable visual:

  1. How much did we sell in total, and how many orders were there?
  2. Which city sells the most?
  3. How do sales trend by month?
  4. Which products sell best in each city (with the numbers)?
  5. Let users choose a date range and a category.
See solution
  1. Two cards at the top left: Total Sales and Number of Orders.
  2. Horizontal bars: Sales by City, sorted, with data labels.
  3. Line chart: Sales by Month, with an average line from the Analytics pane.
  4. Matrix: Product on rows, City on columns, Sales as values, with a background color scale.
  5. Slicers at the top right: Date (Between style) and Category (dropdown).

If you'd rather the cards always show the total for the chosen period, use Edit interactions so they don't filter when someone clicks the line chart.

Summary

  • Choose the visual by the question: how much, who, what's the trend, what's it made of, where, what are the details.
  • Sorted bars to compare, lines for time, a matrix for detail, cards for headline numbers.
  • Filter at the visual, page or report level, and use slicers so users can explore.
  • Control interactions and format with clear titles, little clutter and purposeful color.
  • Bar axes start at zero; with nearly equal parts, like Summit's regions, sorted bars reveal what a pie hides.

Next lesson: How to Design a Professional Power BI Report. You'll design your first complete report, with a page structure, navigation and a professional theme.

🎓 Save your progress and get a free certificate

Create your account in 30 seconds, mark your lessons and download your certificate at the end.

Enroll for free