Skip to Top Main Navigation Skip to Left Navigation Skip to Content Area Skip to Footer
Texas Department of Insurance
Topics:   A B C D E F G H I J K L M N O P Q R S T U V W X Y Z All

Style guide for data visualization tools

Use this guidance when creating dashboards, Power BI reports, maps, charts, tables, and public data displays. The examples follow TDI web style guide patterns while inheriting the normal page scale used by Cascade content pages.

Choose the right visual

Start with the question the user needs to answer. The decision flow below updates the recommendation, visual preview, and accessibility notes based on the selected options.

Data needs
Tip: Use the table option whenever users need exact values or a screen-reader-friendly equivalent.
Map pattern
LowerHigher

Use a map only when geography is the main insight. Pair it with a table.

Bar chart pattern

Use sorted horizontal bars for rankings and long category names.

Line chart pattern

Use line charts for ordered time periods and trends.

Table pattern
County Claims Paid losses
Harris 1,245 $12.4M
Dallas 810 $7.2M
Travis 520 $4.8M

Use tables for precision, lookup, sorting, filtering, and accessible alternatives.

Pie or donut pattern

Use only for simple part-to-whole messages with a few categories.

Scatter plot pattern

Use for correlation, outliers, and relationships between numeric measures.

KPI card pattern

Use KPI cards for a small set of headline measures.

Status pattern

Compliant Inspection current

Expiring soon Insurance expires within 30 days

Expired Action needed

Status uses both color and text, not color alone.

TDI style guide foundation

Use TDI’s website style guide as the base for public data displays. Use Raleway Bold for headings and Open Sans for body text. Use TDI blue for primary emphasis and accessible green, orange, and red only when they support meaning.

Source: TDI style guide

Color roles

TDI blue#124c95Titles, key series, selected states
Green#4e833cPositive status, success, secondary heading
Orange#ce470dWarning, attention, exception
Red#91191cError, severe risk, critical issue
Text#232323Body text, labels, values
Muted gray#5f6a72Captions, notes, metadata
Light gray#eef0f1Panels, filter areas, backgrounds
White#ffffffCards, tables, chart backgrounds

Typography scale

Use point sizes for Power BI themes, design specs, and quick implementation references. Use the same scale consistently across dashboard titles, visual titles, filters, tables, and source notes.

Element Font size to use Recommended implementation Use
Dashboard title 24 pt Use the page <h1> or dashboard title style Report title, public page title, main topic
Section heading 18 pt Use <h2> for major sections Major content sections
Visual title 14 pt Use <h3> or a .visual-title class Chart, map, table, or KPI group title
KPI number 24 pt Use a .kpi-number or card-value style Headline number only
Table, filter, label text 12 pt Use the standard body, table, slicer, and form-label style Values, slicers, labels, table body
Source note 10 pt Use a .source-note, caption, or small-text style Data source, refresh date, definitions

Use a clear dashboard hierarchy

Dashboards should answer the main question first, then let users explore. Keep filters visible, make the selected state obvious, and provide exact data below the main visual.

Title, geography, reporting period, refresh date
Filter/status row: Year, county, coverage type, reset filters, showing all counties
KPI 1
Paid losses
KPI 2
Claims
KPI 3
Average loss
KPI 4
Counties
KPI 5
Top county
KPI 6
Period
Primary visual: map, bar chart, or line chart
Selected summary, definitions, instructions
Data table with exact values
Source, notes, refresh date, limitations

1. Summary first

Use a clear title, filter status, and KPI cards before the user reaches detailed charts or tables.

2. Explore next

Use the main chart or map for interaction, then show a summary of selected counties, categories, or dates.

3. Details last

Use a table for exact values, sorting, export, and screen-reader access to the same information.

Recommended visual patterns

Maps

Use maps when geography is the insight

  • County-level differences.
  • Regional concentration.
  • Geographic gaps in service.
  • Storm, claim, complaint, inspection, or injury distribution.
LowerHigher

Do

Pair the map with a table, legend, selected-county summary, and clear tooltip values.

Avoid

Do not use a map just because the data includes counties. Use a bar chart when ranking is the main task.

Tables

Use tables for exact values, lookup, sorting, filtering, export, and accessible alternatives.

County Claims Paid losses Average loss
Harris 1,245 $12.4M $9,960
Dallas 810 $7.2M $8,889
Travis 520 $4.8M $9,231
  • Use plain-language column names.
  • Right-align numeric values.
  • Use commas, dollar signs, percentages, and units.
  • Include source and refresh date.
Bar charts

Use bar charts for comparing and ranking categories.

  • Sort highest to lowest unless there is a natural order.
  • Use horizontal bars for long names.
  • Start the value axis at zero.
  • Label values directly when possible.
Line charts

Use line charts for trends over time.

  • Use consistent time intervals.
  • Keep the x-axis chronological.
  • Use no more than 3–4 lines unless users can filter.
  • Do not use line charts for unrelated categories.
Pie and donut charts

Use pie or donut charts rarely and only for simple part-to-whole messages.

Use only when

  • There are 2–5 categories.
  • Values add to a meaningful 100%.
  • The message is share, not ranking.
  • Differences are easy to see.
KPI cards

Use KPI cards to summarize a small number of headline measures before users review the main visual.

Power BI dashboard UX best practices

Layout and interaction

  • Make the report fit at 100% browser zoom without cutting off important content.
  • Use a clear title with topic, geography, and timeframe.
  • Place filters in a consistent top row.
  • Show active filters in plain language.
  • Add a reset or clear-filter option.
  • Use a selected-county summary when map selections are available.
  • Provide multi-select instructions, such as “Ctrl+click to select multiple counties.”

Map and table pattern

  • Use a map near the top only when geography is the primary task.
  • Remove unnecessary polygons, decorative layers, and redundant legends.
  • Use a clear legend for low-to-high values.
  • Put a table below the map for exact county-level values.
  • Make KPI cards update based on filters and selections.
  • Use tooltips for supporting details, not essential instructions.

Power BI page example

Homeowners paid losses by county | Texas | FY 2026 | Data refreshed July 2026
Filters: Year, county, coverage type | Status: Showing all counties | Reset filters
Total loss
Claims
Avg. loss
Counties
Top county
Period
Map: paid losses by county
Selected county summary and instructions
County table: County, claims, paid losses, average loss

Make visualizations accessible by design

Every meaningful visual should have an accessible name or alt text, a clear title, sufficient contrast, keyboard access, and a text or table alternative when users need the underlying data.

Color and contrast

  • Use at least 4.5:1 contrast for normal text.
  • Use at least 3:1 contrast for large text and meaningful graphics.
  • Do not use color alone for status or category.
  • Add labels, text, icons, or patterns.

Keyboard and screen readers

  • Set logical tab order in Power BI.
  • Remove decorative objects from tab order.
  • Use clear visual titles.
  • Test keyboard navigation and focus order.

Alternative access

  • Provide tables for exact values.
  • Use Power BI “show as table” access where available.
  • Include summaries for key trends.
  • Do not put essential content only in tooltips.

Alt text examples

Map alt text example

Map showing paid homeowners losses by Texas county for FY 2026. Darker counties represent higher paid losses. Harris County has the highest selected paid loss at $12.4 million. A table below the map provides exact county values.

Bar chart alt text example

Bar chart showing the top 10 counties by paid homeowners losses. Harris County is highest, followed by Dallas and Travis. Exact values are available in the county table below.

KPI card alt text example

Total paid losses for the current selection are $12.4 million. This value updates when county, year, or coverage filters are changed.

Approval checklist

Content

  • The title explains the topic, geography, and timeframe.
  • The data source is visible.
  • The refresh date is visible.
  • Definitions are included where needed.
  • Acronyms are explained.

UX

  • The most important information appears first.
  • Filters are easy to find.
  • Active filters are visible.
  • Reset or clear-filter option is available.
  • Tables provide exact values.

Accessibility

  • Visuals have alt text.
  • Tab order is logical.
  • Color is not the only indicator.
  • Data table alternatives are available.
  • The report works in high contrast mode.

Summary standard

TDI data visualizations should be clear, accessible, and task-focused. Use maps for geography, tables for exact values, bar charts for comparison, line charts for trends, pie charts only for simple part-to-whole data, and KPI cards for headline numbers. Every dashboard should include clear titles, visible filter status, accessible alternatives, readable labels, source information, and TDI-consistent branding.

For more information, contact:

Last updated: 9/2/2026