Going Deeper›Extra· 40 min read

More Visuals: Map, Matrix, Scatter & AI Insights

Beyond bar, line and pie — maps for geography, a matrix for cross-tabs, scatter for relationships, and AI visuals that explain your data.

What you will learn

Pick maps, matrix and scatter for the jobs they suit
Add drill-down and conditional formatting to make visuals explorable
Use the Key Influencers and Decomposition Tree AI visuals

The right visual for the right question

Bar, line and pie cover a lot, but some questions need a different shape. Geography wants a map; a two-way breakdown wants a matrix; “does X relate to Y?” wants a scatter. And Power BI ships AI visuals that find patterns for you. Here is the menu:

Visual	Best for	Example question
Map	Anything geographic	Which cities sell the most?
Matrix	A grid: rows × columns of numbers	Sales by Region (rows) × Product (columns)
Scatter	Relationship between two numbers	Does ad spend relate to sales?
Gauge	Progress toward a target	Are we hitting the 50k goal?
Treemap	Parts of a whole, many categories	Share of sales by product, lots of products

A worked map

To plot sales by city, click the Map visual, drag City into Location and Total Sales into Bubble size. Power BI geocodes the city names (works out their latitude/longitude) and draws a bubble at each one, sized by sales.

A bubble map: each city is a dot sized by its sales

Map:  Location = City,  Bubble size = Total Sales

   Delhi   (big bubble)     12,100
   Pune    (small bubble)      900
   Jaipur  (medium bubble)   1,650

Note: Output: A map of India shows three bubbles — a large one over Delhi, a medium one over Jaipur and a small one over Pune — instantly revealing where the money is geographically. A bar chart could show the same totals, but a map adds the where, which is the whole point of geographic data.

A worked matrix

A matrix is a table with categories down the side and across the top, with a measure in the cells — like a pivot table. Drag Region to Rows, Product to Columns, and Total Sales to Values:

Region ↓ / Product →	Keyboard	Mouse	Monitor
North	1,200	0	8,900
South	0	450	0
East	1,200	0	0

Each cell is the total where that row and that column meet — a fast way to compare two categories at once, with row and column subtotals added automatically.

Drill-down: one visual, many levels

When you put a hierarchy on an axis — Year → Quarter → Month → Day, or Country → State → City — the visual gains a little drill arrow. Click it and the chart zooms from years into quarters into months, all in one visual.

Drag a date field onto a chart’s axis; Power BI auto-creates a Year → Quarter → Month → Day hierarchy.
Click the down-arrow (drill down) icon on the visual to turn drill mode on.
Click a bar (e.g. 2026) to drop into its quarters, then a quarter to see its months.

Note: Output: The chart starts showing four yearly bars. Click 2026 and it redraws as that year’s four quarters; click Q1 and it shows January, February, March. One visual now explores three levels of detail instead of three separate charts.

Conditional formatting: colour that carries meaning

Conditional formatting colours numbers by their value so good and bad jump out. In a table, click a field → Conditional formatting → Background color, and set a rule (e.g. red for low, green for high). A sales table then shades big numbers green and small ones red automatically — readers spot the story without reading every figure.

AI visuals: let Power BI find the pattern

Two built-in AI visuals analyse data for you, no coding needed:

Key Influencers: you pick an outcome (e.g. “high sales”) and a few factors; it ranks which factors most drive that outcome — “when Region is North, sales are 3× more likely to be high”.
Decomposition Tree: you pick a number (Total Sales) and it lets you break it down by any field, click into the biggest part, break that down again — an interactive “why is this number what it is?” explorer.

The Decomposition Tree breaks a number down, biggest part first

Decomposition Tree:  Total Sales  11,750
  +- North   10,100
  |    +- Monitor   8,900   <- click to expand the biggest driver
  |    +- Keyboard  1,200
  +- East     1,200
  +- South      450

Note: Output: Starting from 11,750, the tree splits into regions; North is largest, so you expand it and see Monitor (8,900) is the single biggest contributor. In a few clicks the AI visual answered “what is driving our sales?” — Monitor in the North — without writing any DAX.

Tip: Match the visual to the question: where → map, two categories at once → matrix, does A relate to B → scatter, what is driving this number → Decomposition Tree or Key Influencers.

Watch out: Maps need clean, recognisable place names. “Delhi” geocodes fine, but a typo or an ambiguous name (a town that exists in several countries) can plot in the wrong spot — set the field’s Data category to City/Country to help Power BI place it correctly.

Q. Which built-in AI visual lets you start from a number and repeatedly break it down by any field to find what drives it?

Answer: The Decomposition Tree starts from a measure (e.g. Total Sales) and lets you expand it by field after field, drilling into the biggest contributor each time — an interactive way to find drivers without code.

✍️ Practice

Build a map of Total Sales by City and a matrix of Region (rows) × Product (columns).
Add a date hierarchy to a bar chart and drill from year down to month.

🏠 Homework

Add a Key Influencers or Decomposition Tree visual to a dataset and write two sentences on what it revealed about the drivers of a number.