TL;DR
Use small multiples in the new card visual in Power BI when you compare the same KPIs across regions, brands or teams. Add your measures to Values, drop the split field into Categories and arrange the grid under Multi-category layout. Microsoft folded the preview’s small multiples option into the Categories well, and the visual became generally available in the November 2025 Desktop update.
Fit to space caps the category grid at four rows, so turn Fit to space off in Multi-category layout and set the rows you need. Add reference labels for target and variance so every number arrives with its context. For trends over time, use a chart’s small multiples instead, and for many dimensions at once, use a matrix.
Key Takeaways
The Card icon in current Power BI Desktop creates the new card, which replaces both the legacy Card and the Multi-row card.
Small multiples in the new card now work through the Categories well: one field there repeats every measure for each of its values.
Multi-category layout arranges the category sections; Fit to space stops at four rows, and Fixed size adds scroll bars instead.
Reference labels, conditional formatting and images turn a grid of numbers into a grid of answers.
Use the card for the same KPIs across groups; use chart small multiples for trends over time and a matrix for many dimensions.
Reports built during the preview can shift after the November 2025 update, so recheck padding, rows and reference labels.
Watch on YouTube
Boosting Financial Efficiency with Power BI | Advanced Data Analytics
Kanerika shows how Power BI and advanced analytics lift financial reporting efficiency, the KPI-heavy setting where a well-built card grid earns its place on the page.
Twelve Regions, Thirty-Six Boxes, One Monday Meeting
Picture a regional sales lead who opens the same report every Monday morning. It covers twelve regions, and each region has three cards of its own: revenue, target and variance. That is 36 separate boxes to align, format and re-point.
Then the company splits two regions in half. Someone now has to build six more cards by hand and copy the formatting across. After that, the whole page must be squeezed to make room before the next review.
The new card visual in Power BI removes that chore. One visual holds all three measures, while one field in its Categories well repeats them for every region in the data. When a fourteenth region appears in the table, the grid simply gains another section.
This guide walks through that setup end to end, using a six-region sample dataset, so you can copy each step into your own report.
What Is the New Card Visual in Power BI?
Think of the new card as a container that holds a whole set of numbers. One visual can hold several measures, and each measure then becomes its own card inside the container. You can then add images, reference labels for context and a category field that repeats the cards for every value of that field.
Microsoft’s card visual documentation describes the result plainly: it shows a measure value with optional callout images, reference labels and supporting details. So the card stops being a decoration next to a chart and becomes a compact KPI panel.
There is a performance angle too. Microsoft notes that combining multiple measures and reference labels in one visual reduces visual load time and optimizes the queries sent to the semantic model. Fewer visuals on a page therefore means fewer queries to wait for. If you are new to the wider tool, our overview of Power BI covers the basics first.
What It Replaced: Card and Multi-row Card
Before the November 2025 release, Power BI shipped two separate visuals for headline numbers. The single Card showed one measure, while the Multi-row card listed several values as rows of text. The new card replaces both, and Microsoft recommends it for new reports.
Question Legacy Card Multi-row card New card visual
Measures per visual One Several, as rows of text Several, each as its own card Grid of cards split by a field No No, rows of text Yes, through the Categories well Reference labels for context No No Yes, in their own area Images No No Card, callout and category header images Status in current Desktop Legacy, restorable Legacy, restorable Created by the Card icon
Existing reports that use the legacy visuals still render without changes. Nothing breaks on its own, so you can move pages to the new card when you next touch them.
From Preview to General Availability
The new card reached general availability in November 2025, about two and a half years after the June 2023 preview. That date matters because much of what you read online still describes the preview. Microsoft’s Power BI Desktop update archive also records each step.
June 2023 (2.118.286.0): the new card arrives as a preview feature.November 2024 (2.138.782.0): small multiples for the card appear in preview.March 2025 (2.141.1228.0): category enhancements follow, still in preview.November 2025: general availability, with hero images, a collage layout and display order controls.February 2026: default callouts rise from 5 to 10, and selecting a category name filters other visuals.April 2026 (2.153.910.0): Fixed size arrives, Autogrid becomes Fit to space, and multiple category fields concatenate.June 2026 (2.155.756.0): hover-state image formatting becomes generally available.August 2026: image files stored in OneLake can serve as card image sources.
The practical lesson is simple. Any tutorial that starts with a Preview features checkbox was written before November 2025, and its layout advice may no longer match your Desktop.
Small Multiples or Categories?
People still search for “small multiples” in the card, and that wording made sense during the preview. Today, though, the same idea lives in the Categories field well, and its layout sits under Multi-category layout in the format pane. The words changed, but the behavior you want did not.
Preview-era wording Current wording (since November 2025)
Small multiples field Categories field well Small multiples layout Multi-category layout Rows or columns for a grid, vertical or horizontal arrangement Only the grid setting carries over; vertical or horizontal settings reset to 10 columns and 10 rows Autogrid Fit to space (renamed April 2026) Preview features checkbox Not needed; the Card icon creates the new card
So when this guide says small multiples, read it as a card visual with a field in Categories. The two terms describe the same grid.
How to Add the New Card Visual in Power BI Desktop
Adding the visual takes a minute when your Desktop is current. Power BI Desktop is a free download , so updating costs nothing but the install time.
Open your report in Power BI Desktop, November 2025 or later.
In the Visualizations pane, select the Build visual tab.
Select the Card icon from the visual gallery.
Drag a measure or a data column into the Values field well.
Finally, resize the visual so it leaves room for more cards.
That gives you a single card. When you add more values, more cards appear inside the same visual.
On a Desktop Build Older Than November 2025
Update Desktop before you build anything. An older build still shows the preview version, and a report published from it gets upgraded to the generally available card in the service. That upgrade is why a card can look different online than it did on your laptop. The old Preview features toggle is history now, so skip any guide that sends you there.
Where you publish matters as well. Power BI Report Server uses its own build of Desktop, released on a slower schedule than the monthly Desktop updates. Check that build’s release notes for the card visual before you standardize a report on it.
How to Bring Back the Legacy Card and Multi-row Card
Some teams keep the old visuals for backward compatibility, and Microsoft leaves a path open for them.
In the Visualizations pane, select the three dots (…).
Then select Restore default visuals.
Find the legacy Card and Multi-row card in the unpinned area of the pane, where custom visuals usually sit.
Use them to maintain old pages rather than to start new ones.
How to Build Small Multiples in the New Card Visual, Step by Step
Building one is the fastest way to learn this grid. Each step below uses the same regional sales example, with three measures, one Region field and six regions in the sample data.
The payoff is easy to count. Three measures across six regions would need 18 legacy cards, each placed and formatted by hand. Here they come from one visual that you format once.
Step 1: Prepare the KPI Measures
Start in the model before you touch the canvas. Cards behave best when measures sit on a clean star schema, with a Sales fact table and a Region dimension. Our comparison of star schema vs snowflake schema also explains why flat dimensions suit Power BI. Write three explicit measures so every card shows a deliberate number.
Revenue = SUM ( Sales[Revenue] )
Target = SUM ( Targets[Target] )
Variance % =
DIVIDE ( [Revenue] - [Target], [Target] )DIVIDE returns a blank instead of an error when a region has no target, which keeps the grid clean. For more modeling habits that keep measures fast, see our Power BI data modeling best practices .
Kanerika Service
Need KPI Measures That Hold Up in a Card Grid?
Kanerika builds the star-schema models and DAX measures behind small multiples card grids, then tests each card with the roles that will open the report.
Explore Power BI Services Step 2: Add Measures to Values
Each measure you drop into Values becomes its own card. Order matters, because the first measure usually reads as the headline.
Select the card visual on the canvas.
Drag Revenue into the Values field well.
Leave Target and Variance % out of Values for now; they become reference labels in Step 5.
Since the February 2026 update, the visual shows up to 10 callouts by default, up from 5. Keep the count lower than that on a busy page.
Step 3: Add a Field to Categories
This is the small multiples moment. Drag Region from the Data pane into the Categories field well. The visual then builds a separate section for each region, and each section shows Revenue sliced to that region.
You can also add a second field. Since April 2026, multiple category fields concatenate in the section header, so Region plus Product line produces headers such as “North, Bikes”. Use that sparingly, because two fields multiply the number of sections quickly.
Step 4: Set the Multi-category Layout
Now decide how the sections sit on the page. Open the Format visual icon, choose the Visual tab and expand Multi-category layout. Set the number of rows and columns for your grid.
One limit catches people out. With Fit to space on, the grid stops at four rows, so a 6 by 1 stack will not appear as you set it. Turn Fit to space off to show more rows. Alternatively, switch on Fixed size in the multi-card layout, which gives every card exact pixel dimensions and adds scroll bars.
For six regions, a 2 by 3 grid with Fit to space on is the tidy choice.
Step 5: Add Reference Labels for Target and Variance
A revenue figure alone does not say whether a region is winning. Reference labels supply that answer underneath each callout.
In the Format visual pane, expand Reference labels.
In Apply settings to, select Revenue.
Drag Target into the Add label data field.
Finally, turn on Detail and drag Variance % into the Details field.
Reference labels take half the card by default. If they crowd the callout, set Apply settings to All under Cards, then raise the Callout size % in Layout until the balance looks right.
Step 6: Apply Conditional Formatting
Color should point at the regions that need a conversation. The card supports conditional formatting on font color, background color and border color, using the same rule types as other Power BI visuals.
A practical rule for this example is a red border when Variance % falls below zero. The sales lead then sees the trailing regions before reading a single number. Keep it to one color signal per card, or the grid starts to shout.
Step 7: Test Category Selection and Interactions
Category headers are more than labels. Selecting one filters the other visuals on the page, and the remaining sections dim to show the current choice.
Hold Ctrl and select more headers to filter by several regions at once.
Select a header again when you want to clear it.
Use Edit interactions on the Format menu to choose which visuals respond.
Also test this with a slicer on the same page. A text slicer visual in Power BI lets users type a region name and watch the card grid narrow to match.
Sample Data and Expected Result
Here is a small sample dataset you can type into a table to follow along. The figures are illustrative, not real company data, and Variance % is (Revenue minus Target) divided by Target.
Region Revenue ($) Target ($) Variance %
North 1,240,000 1,200,000 3.3% South 980,000 1,050,000 -6.7% East 1,115,000 1,100,000 1.4% West 870,000 900,000 -3.3% Central 1,060,000 1,000,000 6.0% Northwest 640,000 700,000 -8.6%
With the steps above, the visual shows six sections in a 2 by 3 grid. Each one carries Revenue as the callout, Target as a reference label and Variance % as the detail. South, West and Northwest pick up the red border, so three of six regions stand out at a glance.
Talk to Kanerika
Want This KPI Grid Built on Your Own Data?
Bring one report page and its KPI list. Kanerika will model the measures, set up the card grid by region or brand, and test it with the roles that will use it.
Talk to a Power BI Expert → Layout Options: Multi-card, Multi-category, Fixed Size and Fit to Space
Layout is where most confusion starts, because three different controls arrange three different things. Once you see which control owns which level, the format pane stops feeling like a maze.
Think of it as nesting boxes. Cards Layout arranges the parts inside one card, Multi-card layout arranges the cards inside a section, and Multi-category layout arranges the sections inside the visual.
Control What it arranges When to change it
Cards > Layout Callout, image and reference labels inside each card, including the collage layout and display order When labels crowd the number or an image needs more room Multi-card layout The cards (one per measure) inside the visual or inside each category section When you show several measures per region Multi-category layout The category sections, such as one per region When you use a field in Categories Fit to space Cards or sections grow or shrink to fill the container; caps the category grid at four rows When you have four rows or fewer and want no empty slots Fixed size Exact pixel size per card, with scroll bars when space runs out When readability matters more than seeing every card at once
Fixed size and Fit to space are mutually exclusive. Switching on Fixed size disables Fit to space, because the pixel dimensions now control the layout. With Fit to space off and no Fixed size, the visual reserves room for the number of cards you set, even when fewer exist.
During the preview you could set rows or columns for a grid, vertical or horizontal arrangement. Since November 2025 only the grid setting carries over, and vertical or horizontal settings reset to 10 columns and 10 rows. If an older report suddenly shows a stretched grid, set its rows and columns again.
Formatting the Cards: Callouts, Images, Reference Labels and Headers
Formatting decides whether the grid gets read or skimmed past. The aim is a card a manager can understand in two seconds, from across the meeting room.
Start with the callout value, because it is the largest text on the card. Then add context with reference labels, and only after that reach for images and colors. If you are planning a whole page around these cards, our guide to Power BI dashboard development covers layout at page level.
Display Units: 12K or 11,732
By default, the callout uses Display units set to Auto, which abbreviates large numbers with K, M or B. Microsoft’s own example: 11,732 appears as 12K. That suits an executive grid but frustrates a finance reviewer who needs the exact figure.
Select the card visual and open the Format visual icon.
Under the Visual tab, expand Callout values, then Values.
Finally, set Display units to None.
If a custom format string still shows a K, M or B suffix, the Display units setting is the cause, not the format string.
Images and Hover States
Images can make a card grid easier to scan, as long as each one carries meaning. The card supports three image types, and each comes from one of three sources.
Image type Where you set it Good use
Image (hero image) Image section of the format pane Product photos or a small trend chart beside the value Callout image Callout > Image A small icon next to the number Category header background image Category header section A brand or region marker on each section
Any of the three can come from an upload, an image URL or a data column. Since the August 2026 update , that URL can also point to an image file in OneLake. If your images sit in a Fabric lakehouse, our explainer on data lake vs lakehouse shows where that storage fits.
The Image section also has a States dropdown with Default and Hover settings. A hover state can swap the image source, change the fit or alter the background color when a user points at a card. Sometimes every card shows the same image after you add a category. In that case, check Apply settings to, because a setting of All sends one image to every card.
Category Headers, Padding and Theme Defaults
Category headers deserve the same care as the numbers. Give each header a readable font size, and add a header image only when it helps people find their region faster.
Padding is the other common surprise. The default padding is 12px since general availability, so cards built during the preview can look roomier than before. Search for Padding in the format pane to adjust it. Microsoft also documents a theme file entry that sets padding once for every card in a report, using its report theme format.
{
"name": "CardVisualGA",
"visualStyles": {
"cardVisual": {
"*": {
"*": [{
"$id": "default",
"paddingUniform": 0
}]
}
}
}
}A theme keeps a 40-card report consistent without formatting each visual by hand. Set the value your design system uses rather than copying zero blindly.
Fixing Preview-Era Reports After the November 2025 Update
If you built cards during the preview, the November 2025 release probably moved something. None of these changes are bugs in your report; they come from new defaults that Microsoft lists in its card documentation.
The table maps each symptom to its cause and the fix, so you can work through an older report one page at a time.
Problem you see Cause Fix
Columns or rows reset to 10 Only the grid arrangement was kept; vertical and horizontal reset to 10 columns and 10 rows Set rows and columns again in Desktop (November 2025 or later), then republish Cards take up more space Padding grew because the default is now 12px Search Padding in the format pane, then try individual padding or a theme Reference label covers the callout Reference labels have their own area, 50% of the card by default Cards > Apply settings to All > Layout, then adjust Callout size % Empty reference label area or a new background Once one card uses the area, the others keep it; a default background applies Open Reference labels layout, then toggle Background off Image looks zoomed in or its settings moved Callout images moved to Callout > Image; preview sizing bugs were corrected Set image fit to Fit or Center, then check the size controls in Callout > Image Card looks different in the service than in Desktop Desktop older than November 2025, upgraded on publish Update Desktop; otherwise, edit the report on the web after publishing
Work through the rows in order on a copy of the report. Rows and padding fixes usually solve most of the visible drift before you touch images.
Checklist
Enterprise Power BI Checklist
Check performance, governance, security, licensing and Microsoft Fabric readiness before your next Power BI report goes live.
Get the Checklist → Performance, Security and Accessibility Checks
A card grid often lands on the most-viewed page of a report, so test it the way users will meet it. A field in Categories multiplies the sections the visual has to render, and row-level security changes which sections each user sees at all. Both only show up with real data volumes and real roles.
Before you publish, run these five checks.
Replace several single cards with one card visual where the measures belong together.
Test with real roles if the model uses Power BI row-level security , so each user sees only their regions.
Pair every color rule with a text signal, such as a negative Variance %, so meaning never depends on color alone, as WCAG’s use of color guidance requires.
Also check the grid in the mobile layout, since four columns rarely fit there.
Finally, confirm that header selection filters only the visuals you intend.
Row-level security matters most here, because a category grid can reveal which regions exist even before anyone reads a number.
How Kanerika Builds Power BI KPI Reporting
Large KPI estates are where a well-built card grid pays off, and they are also where it goes wrong fastest. Our Power BI consulting team designs these pages model first. Every measure gets a named owner and a definition, and every page is tested with the roles that will open it.
One engagement shows the scale involved. A multi-brand CPG company ran five separate financial applications after a series of acquisitions. Kanerika integrated that data with Informatica and built Power BI reporting for 70+ KPIs , including financial statements. The client recorded a 60% increase in operational performance, a 40% decrease in market response time and a 100% improvement in data integration.
A 70-KPI estate across brands and business units is exactly where card grids help. One card visual per KPI group, split by brand through Categories, can then replace dozens of single cards. When a card flags a breach, its owner still needs an alert. Our comparison of Power Apps vs Power Automate shows which tool runs that flow.
Case Study
Power BI Reporting for 70+ KPIs Across CPG Brands
Kanerika unified five financial applications with Informatica and built Power BI reporting for 70+ KPIs, including financial statements, for a multi-brand CPG company.
Read the Case Study → When to Use Small Multiples in the New Card Visual (and When Not To)
So, should your next KPI page use small multiples in the new card visual? Use it when the question is “how does the same number compare across groups”, and pick another visual when the question changes shape.
Your need Best choice Why
Same KPIs by region New card with Region in Categories One visual, so one format and a section per region Same KPIs by brand or product line New card with Brand in Categories Header images can also carry brand marks Same KPIs by channel or team New card with Categories, Fit to space on Few groups, so they fit within the four-row cap Same KPIs by business unit, 20+ units New card, Fit to space off, rows set to fit (Fixed size on the cards if you want scroll bars) Setting the rows yourself shows every unit past the four-row cap Trend over time per group Not the card: use small multiples on a line chart Charts show shape over time, while cards show one value Many dimensions at once Not the card: use a matrix Rows and columns can also cross two or more fields One number against one target A single card with reference labels, or a KPI visual No groups to compare, so no Categories needed
Pair the Card Grid With Chart Small Multiples
Chart small multiples are the natural partner here. Microsoft supports small multiples on bar, column, line and area charts . So a page can pair a card grid of current values with a line chart grid of trends.
Take the regional sales page from this guide. Put the card grid of this month’s revenue by Region at the top, and below it place a line chart with Region in Small multiples. Selecting a region’s header in the card then filters the line chart, so the current number and its trend line up for the same region.
The Verdict on Small Multiples in the New Card
The verdict is short. For the same KPIs across groups, small multiples in the new card visual are the right tool, and the setup takes four moves. Put the measures in Values and one field in Categories, then choose a layout that fits the grid. Finally, add reference labels so each number arrives with its context. That is how you use small multiples in the new card visual: one visual, four settings and every group on one page.
Frequently Asked Questions
Is the new card visual replacing the old one in Power BI?
Yes. Since the November 2025 Power BI Desktop update, the Card icon creates the new card visual, which replaces both the legacy Card and the Multi-row card. Existing reports that use the legacy visuals keep rendering without changes, and you can restore the old visuals from the Visualizations pane when you need to maintain older pages.
How do I enable the new card visual in Power BI?
On Power BI Desktop from November 2025 onward, there is nothing to enable. The new card is the default Card visual, so you select the Card icon on the Build visual tab and drag a measure into Values. On an older build, update Desktop first, because the preview checkbox from earlier tutorials no longer applies.
Can I apply conditional formatting to the new card visual?
Yes. The new card supports conditional formatting on font color, background color and border color, using the same rule types as other Power BI visuals. A common rule turns a card border red when variance to target falls below zero. Pair the color with a visible number or label, so the meaning never depends on color alone.
Can I add more than one field to Categories in the card visual?
Yes. Since the April 2026 update, you can place more than one field in the Categories well, and their values concatenate in each section header, such as a region and a product line together. Each extra field multiplies the number of sections, so check the grid size and consider Fixed size with scroll bars for long lists.
Is the new card visual free to use?
The card visual is part of Power BI itself, not a paid add-on or marketplace visual. Power BI Desktop, where you build it, is a free download from Microsoft. Sharing reports with colleagues through the Power BI service depends on your organization’s licensing, so check your plan before you publish a card-based report widely.
Can I use small multiples in other visuals besides the card visual?
Yes. Power BI supports small multiples on bar, column, line and area charts, where one field splits a chart into a grid of matching charts. Use chart small multiples to show trends over time for each group, and use the card visual with Categories when you compare current KPI values across regions, brands or teams.
Why don't I see the small multiples option in my card visual?
The option now works through a different name. In the generally available card visual, small multiples come from the Categories field well on the Build visual tab, and the grid settings sit under Multi-category layout in the format pane. If no Categories well appears, check that you are using the new Card and not a restored legacy visual.
Can I show more than one measure in each small multiple card?
Yes. Every measure you add to Values becomes its own card inside each category section, so three measures across six regions produce eighteen cards from one visual. You can also add reference labels and detail values under each callout, which is often cleaner than adding more callouts when the extra numbers only provide context.
How does Power BI handle too many cards in small multiples?
With Fit to space on, the category grid stops at four rows, so extra sections need another setting. Turn Fit to space off to set more rows, or switch on Fixed size to give each card exact pixel dimensions with scroll bars. The default callout limit also rose from five to ten in the February 2026 update.
How do I get the old Card and Multi-row card back in Power BI?
Open the Visualizations pane, select the three dots and choose Restore default visuals. The legacy Card and Multi-row card then appear in the unpinned area of the pane, where custom visuals usually sit. Microsoft keeps them for backward compatibility, so use them to maintain older pages and build new KPI pages with the new card.
Why does my card look different in the Power BI service than in Desktop?
The usual cause is an older Desktop. If you built the report on a version from before November 2025, the service upgrades the card to the generally available version when you publish, and new defaults such as 12px padding apply. Update Power BI Desktop, or edit the report on the web after publishing to adjust the style.
How do I show the full number instead of 12K in a Power BI card?
Select the card, open the Format visual icon and expand Callout values, then Values. Set Display units to None. The default setting, Auto, abbreviates large numbers with K, M or B, so 11,732 shows as 12K. If a custom format string still shows a suffix, the Display units setting is the cause rather than the format string.
How do I add a target and variance to a Power BI card?
Create Target and Variance measures first, then open Reference labels in the format pane. Choose the callout in Apply settings to, drag Target into the Add label field, turn on Detail and drag Variance into Details. Reference labels take half the card by default, so lower the callout size in Cards Layout if the labels crowd the number.
