Understanding Charts
Chartlet.AI generates charts from your dataset profile, column types, and common analysis patterns. Use this guide to review what it creates and decide what belongs in your dashboard.
Compare categories
Bar, column, table, KPI tile
Sales by product, revenue by region, customer counts, or ranked lists.
Show trends
Line, area, sparkline
Daily sales, monthly revenue, traffic, conversion rates, and other time-based metrics.
Show composition
Pie, donut, treemap, stacked charts
Category share, product mix, campaign contribution, or budget allocation.
Find relationships
Scatter, bubble, correlation matrix
Price versus sales, discount versus margin, customer age versus spend, or metric correlations.
Spot distribution and outliers
Histogram, box plot, heatmap
Order sizes, response times, missingness, intensity, and unusual values.
Track funnel or target progress
Funnel, waterfall, gauge
Pipeline stages, revenue bridges, goal progress, or conversion flow.
How Chartlet.AI Chooses Charts
Chartlet.AI looks at column names, detected data types, date fields, numeric measures, categories, and dataset shape. It then suggests charts that are likely to explain the data quickly.
Generated charts are a starting point. You should keep charts that help answer a business question and remove or edit charts that are confusing, repetitive, or not relevant to the audience.
Chart Review Checklist
- Does the title explain the metric and segment?
- Are date ranges and currencies clear?
- Do labels use business terms your audience understands?
- Is the chart showing a real pattern or just noise?
- Should the chart be removed before sharing the dashboard?
Dashboard and Chart Actions
- Review generated charts on the dataset dashboard.
- Add or edit charts when you need a more specific view.
- Arrange dashboard cards so the most important metrics appear first.
- Use chart download actions where available for a quick export.
- Pin useful AI-generated charts back to the dashboard.
When a Chart Looks Wrong
A chart can be misleading when source data has mixed types, missing values, duplicate headers, unusual date formats, or columns with unclear names. Start by checking the Data Explorer, then edit or remove the chart if it does not answer a useful question.