Skip to main content

How to Use the AI Query Helper to Shape Chart Visualizations

Here's how to work with the AI Query Helper to create and modify charts for your data.

Ashley Dehertogh avatar
Written by Ashley Dehertogh
Updated over 3 weeks ago

Getting Started with Charts

When working with data in Insights, visualizations help you understand trends and patterns at a glance. Here's how to work with the AI to create and modify charts for your data.

Requesting a Specific Chart Type

To specify a chart type in your initial query, simply include the chart type in your request:

  • "Show me monthly revenue as a bar chart"

  • "Display occupancy by segment as a pie chart"

  • "Create a line chart showing ADR trends for the past 6 months"

Changing an Existing Chart

If you already have a query and want to change the visualization:

  1. Use clear action words followed by the desired chart type:

    • "Change this to a column chart"

    • "Convert to a stacked bar chart"

    • "Display as a heatmap instead"

    • "Switch to a line chart"

  2. Be specific about what you want to visualize:

    • "Show this data as a stacked column chart with segments"

    • "Change to a heatmap showing occupancy by day of week"

Available Chart Types

Bar & Column Charts

  • Column Grouped: "Show as a grouped column chart" (vertical bars, side-by-side)

  • Column Stacked: "Change to a stacked column chart" (vertical bars, stacked)

  • Column Stacked Percentage: "Convert to a percentage stacked column" (vertical bars, showing proportions)

  • Bar Grouped: "Show as a grouped bar chart" (horizontal bars, side-by-side)

  • Bar Stacked: "Display as a stacked bar chart" (horizontal bars, stacked)

  • Bar Stacked Percentage: "Use a percentage stacked bar chart" (horizontal bars, showing proportions)

Line & Area Charts

  • Line Chart: "Show as a line chart" (good for time series)

  • Line Chart with Colors: "Display as a colored line chart" (multiple lines with distinct colors)

  • Area Stacked: "Convert to a stacked area chart" (filled areas, stacked)

  • Area Stacked Percentage: "Change to a percentage stacked area chart" (shows proportions over time)

Other Visualizations

  • Pie Chart: "Show as a pie chart" (composition of a whole)

  • Map: "Display as a map" (geographic data)

  • Region Map: "Use a region map" (data by geographic regions)

  • Heatmap: "Convert to a heatmap" (color intensity showing values across two dimensions)

  • Table: "Show as a table" (detailed numeric data)

  • Point Size/Color: "Display as a scatter plot" (relationships between variables)

Chart Selection Recommendations

Different data types work best with specific chart types:

Data Type

Recommended Charts

Example Query

Time series

Line chart, Area chart

"Show ADR trends by month as a line chart"

Categorical comparisons

Bar chart, Column chart

"Display revenue by segment as a bar chart"

Part-to-whole relationships

Pie chart, Stacked charts

"Show revenue breakdown by room type as a pie chart"

Geographic data

Map, Region map

"Show occupancy by property location as a map"

Multi-dimensional analysis

Heatmap

"Display occupancy by day of week and month as a heatmap"

Showing correlations

Scatter plot

"Show the relationship between ADR and occupancy as a scatter plot"

Automatic Chart Selection

If you don't specify a chart type, the AI will choose an appropriate visualization based on:

  • The number and types of dimensions and measures in your query

  • The presence of time-series data

  • The overall data structure and intended analysis

You can always ask to change the chart type after seeing the initial visualization.

Limitations of Chart Customization

While the AI can help you select appropriate chart types, there are some limitations to be aware of:

  1. Limited Formatting Options: The AI can only specify the chart type, not detailed formatting such as:

    • Custom colors, fonts, or font sizes

    • Axis formatting (logarithmic scales, custom ranges)

    • Chart titles, legends, or label customizations

    • Custom tooltips or annotations

  2. Data-dependent Constraints:

    • Some chart types only work with specific data structures (e.g., maps require geographical data)

    • Pivot tables may limit certain visualization options

    • Large datasets might render better in some chart types than others

  3. No Preview: The AI cannot show you a preview of the chart before generating it

  4. No Combined Chart Types: Cannot create hybrid visualizations (e.g., combining a line and bar chart)

  5. No Custom Calculations for Charts: Cannot create visualizations based on custom calculations not in the original data

  6. No Direct Chart Editing: For detailed visual customization, you'll need to use Insights' built-in chart editor after the visualization is generated

Best Practices

Choose the right chart for your data:

  1. Time Series: Line charts or area charts for trends over time

  2. Comparisons: Bar/column charts for comparing values across categories

  3. Composition: Pie charts or stacked charts for showing parts of a whole

  4. Distribution: Heatmaps for seeing patterns across multiple dimensions

  5. Relationships: Scatter plots for correlations between variables

Examples

Here are complete examples you can use:

  • "Show me monthly ADR and RevPAR for this year as a line chart"

  • "Display room type revenue breakdown as a pie chart"

  • "Show occupancy by day of week as a heatmap"

  • "Compare property performance as a grouped bar chart"

  • "Show segment revenue as a stacked column chart"

  • "Display month-over-month occupancy growth as a line chart"

By specifying the type of visualization you need, you'll get insights presented in the most meaningful way for your analysis requirements.


πŸ“Œ Need more information? Reach out to our Advisory team via the chat with any feedback on particularly good examples you've used and/or you need any support!

Did this answer your question?