Query chart types

The Query report builder has 12 chart types in five categories: Time Series, Distribution, Composition, Correlation and Metrics. This page covers what each chart type shows, when to use it, and the parameters it needs in the UI and in an MCP prompt. A quick-reference matrix of all 12 is at the end. Pick the chart type from the Chart Type dropdown at the top of the report builder.

Chart Type dropdown open in the report builder, grouping Bar, Line, Stacked and Combo under Time Series, Histogram under Distribution, Pie, Treemap and Waterfall under Composition, Scatter and Heatmap under Correlation, and KPI and Impact under Metrics.

When to use

Use case	Chart type
Compare discrete time buckets	Bar
Spot a trend over continuous time	Line
Show part-to-whole over time	Stacked
Compare two measures on different scales	Combo
See how values are spread	Histogram
Show proportions across a few categories	Pie
Compare relative sizes across many categories	Treemap
Show how components add up to a total	Waterfall
Find a correlation between two metrics	Scatter
Find patterns across two dimensions	Heatmap
Show one headline number	KPI
Measure the effect of an event on a metric	Impact

Common parameters

Every chart type needs these parameters.

Parameter	Required?	Description
Schema	Yes	The data schema (event table) to query
Measure	Yes	At least one measure: a function (Count, Count Distinct, Sum, Avg, Min, Max, Percentile) applied to a field. Some chart types need more than one.
Date Range	Yes	The time window, such as “Last 7 days”, “Last 30 days” or a custom start and end
Filter	Optional	Global conditions applied before aggregation. Operators: equals, not equals, contains, starts with, ends with, greater than, less than, between, exists, is empty and regex matches. Conditions combine with AND or OR.
Cohort	Optional	Limits results to a saved user cohort
Sample %	Optional	Sampling rate for large datasets

Filters and Sampling cover the optional parameters in detail.

Time Series charts

Time Series charts plot data over time. They support the time units Hour, Day, Week, Month, Quarter and Year, plus All Time.

Bar

A Bar chart shows values as vertical bars over time periods. Use it to compare discrete time buckets.

Parameter	Required?	Notes
Measure	Yes	One or more measures
Time Unit	Yes	Hour, Day, Week, Month, Quarter, Year or All Time
Breakdown	Optional	Splits bars by a dimension, such as country or plan
Comparison	Optional	Compares against a prior period, such as the previous 7 days

Bar chart of daily Add to Cart tool executions over 7 days, grouped by Primary Intent into Total, product_inquiry, technical_support and N/A.

Line

A Line chart connects data points over time. Use it to spot trends across continuous time periods.

Parameter	Required?	Notes
Measure	Yes	One or more measures
Time Unit	Yes	Hour, Day, Week, Month, Quarter, Year or All Time
Breakdown	Optional	Adds a line per dimension value
Comparison	Optional	Overlays a prior period

Line chart of the same daily Add to Cart counts, with one line per Primary Intent value.

Stacked

A Stacked chart divides each column into segments by a breakdown dimension. Use it to show part-to-whole relationships over time.

Parameter	Required?	Notes
Measure	Yes	One or more measures
Time Unit	Yes	Hour, Day, Week, Month, Quarter, Year or All Time
Breakdown	Yes	At least one breakdown group. Each segment is one dimension value.
Comparison	Optional	Compares against a prior period

Stacked column chart of daily Add to Cart counts, each column split by Primary Intent, with number and percent toggles and a Hide Others button above the chart.

Combo

A Combo chart overlays bars and lines on the same time axis. Use it to compare two measures with different scales, such as revenue as bars and conversion rate as a line.

Parameter	Required?	Notes
Measure	Yes	Two or more measures recommended. One renders as bars and another as a line.
Time Unit	Yes	Hour, Day, Week, Month, Quarter, Year or All Time
Breakdown	Optional	Splits by a dimension
Comparison	Optional	Compares against a prior period

Set which measure is a line or bars, and which Y-axis it uses, under Measure Assignments in Chart Settings.

Combo chart with Add to Cart counts as lines on the left axis and Place Order counts as bars on the right axis, and Chart Settings open with Measure Assignments set to Line and Bar.

Distribution charts

Histogram

A Histogram shows how a measure’s values spread across buckets. Use it for questions like “How are session durations distributed?”

Parameter	Required?	Notes
Measure	Yes	One measure. Several measures need a formula that combines them.
Time Unit	All Time only	A Histogram aggregates across the whole date range.
Binning	Yes	How values are bucketed: auto range (min, max, bin count), manual ranges or enum buckets
Breakdown	Optional	1 breakdown group at most
Comparison	Not supported
Formula	Conditional	Required with several measures (a compound measure)

Build a Histogram

Set the chart type in Query to Histogram.
Build the measure you want to analyze. Typically this is Sum, a count of events, or a compound measure such as a percentage per user.
Select the subject to analyze. The subject is what gets counted, the thing that performs the measure from step 2.
Choose the count subject on the Y-axis.
Choose the Group By on the X-axis. This sets the bins. Use a pre-built range if it fits, or create a custom one. Base the range on the values you expect from your measure. If a user can do an action at most 10 times a day, the range’s maximum is likely 10.
Apply any filters you need.
Select the date range.
Click Execute.

Group by and binning covers bin ranges in detail.

Composition charts

Pie

A Pie chart shows the proportions of a measure across a few categories. Use it for a quick part-to-whole snapshot.

Parameter	Required?	Notes
Measure	Yes	One or more measures
Time Unit	All Time only	A Pie chart aggregates across the full date range.
Breakdown	Yes	At least one breakdown group. Each slice is one dimension value.
Comparison	Optional	Compares against a prior period

Donut-style Pie chart of Add to Cart counts by Primary Intent, with a total of 996 in the center and slices labeled 82%, 12% and 6%.

Treemap

A Treemap shows proportions as nested rectangles. Use it to compare relative sizes across many categories, especially with hierarchical breakdowns.

Parameter	Required?	Notes
Measure	Yes	One measure. Several measures need a formula.
Time Unit	All Time only	A Treemap aggregates across the full date range.
Breakdown	Yes	At least one breakdown group
Comparison	Not supported
Formula	Conditional	Required with several measures

Treemap of product viewed events grouped by product_id, with larger, darker rectangles for products with more views and a Low to High color scale below.

Waterfall

A Waterfall chart shows how individual components add up to a total. Use it to see which parts contribute most, such as which event types make up the total count.

Parameter	Required?	Notes
Measure	Yes	One measure. Several measures need a formula.
Time Unit	All Time only	A Waterfall chart aggregates across the full date range.
Breakdown	Yes	At least one breakdown group. Each bar is one dimension value’s contribution.
Comparison	Not supported
Formula	Conditional	Required with several measures

Waterfall chart of product viewed events by product_id, with each product's count stacking up to a total bar of 1,496 on the right.

Correlation charts

Scatter

A Scatter chart maps measures to the X-axis, the Y-axis and bubble size. Use it to spot a correlation between two metrics, with a third metric as context.

Parameter	Required?	Notes
Measure	Yes	An X-axis measure and a Y-axis measure. Bubble Size is a third measure.
Time Unit	All Time only	A Scatter chart aggregates across the full date range.
Breakdown	TBD	Adds a label and color per dimension value
Comparison	Not supported
Formula	Conditional	Required when several measures are used as compound inputs

Scatter chart of agent spans with Unique Sessions on the X-axis and average sentiment drift on the Y-axis, each bubble labeled with a tool or agent name, and the builder showing X-Axis, Y-Axis and an optional Bubble Size measure.

Heatmap

A Heatmap plots two dimensions on the X and Y axes, with color intensity showing the measure value. Use it to find patterns across two categorical dimensions, such as activity by day of week and hour.

Parameter	Required?	Notes
Measure	Yes	One measure. Several measures need a formula.
Time Unit	All Time only	A Heatmap aggregates across the full date range.
Breakdown	Yes, exactly 2 groups	The first group maps to the X-axis and the second to the Y-axis. Both need a field.
Comparison	Not supported
Formula	Conditional	Required with several measures

Heatmap of average friction score with Primary Intent values across the X-axis and friction signal types down the Y-axis, darker red cells marking higher scores.

Metrics charts

KPI

A KPI chart shows one large number per measure, with an optional trend. Use it on dashboards and for at-a-glance indicators.

Parameter	Required?	Notes
Measure	Yes	One or more measures. Each renders as its own KPI card.
Time Unit	Yes	Hour, Day, Week, Month, Quarter, Year or All Time
Breakdown	Optional	Groups the KPI
Comparison	Optional	Shows the change against a prior period, such as “+12% vs. last week”

KPI chart showing an average sentiment drift of -0.14, the previous week's value and a 1.40% change, with Compare set to Previous Week.

Impact

An Impact chart measures how a trigger event affects a metric over time. Use it to evaluate the effect of a product change, feature launch or campaign. The question it answers is “What is the impact on Metric A after a user engages with Event A?”

Parameter	Required?	Notes
Measure	Yes	The metric to evaluate the impact on
Trigger Event	Yes	The event that marks the moment whose impact you want to measure
Time Unit	Yes	Hour, Day, Week, Month, Quarter, Year or All Time
Breakdown	Optional	Compares the impact across segments
Comparison	Not supported

Example questions for an Impact chart:

Metric	Trigger event
Total time watched	A content recommendation is shown
Total sessions	The user signs up
Customer lifetime value (LTV)	The first premium purchase

Build an Impact analysis

Set the chart type in Query to Impact.
Build the measure that the event should affect. Typically this is Sum, a count of events, or a compound measure such as a percentage per user.
Apply any filters or breakdowns.
Set the date range.
Select the trigger event: the event you expect to affect the measure from step 2, such as Campaign Impression, Content Recommended or Signed In.
Click Execute.

Impact chart with Add to Cart as the trigger event for the User subject and Place Order count as the measure, plotted from Day -6 to Day 6 around a Day 0 marker.

Read the results

Part of the chart	What it shows
Day 0 marker	The first time in the date range that each user triggered the event. Kubit aligns every user on this day.
Negative and positive day counters	The days before and after Day 0. Each user’s Day 0 can be a different calendar date, so Kubit normalizes them.
The line	The measure from step 2, so you can see whether it changed after the trigger event

In the example above, the Place Order count rises from Day 0, which shows that Add to Cart had a positive effect on order placement in that period.

Quick reference

Chart type	Time granularity	Breakdown	Comparison	Date Kind	Special requirements
Bar	Any	Optional	Yes	Yes	None
Line	Any	Optional	Yes	Yes	None
Stacked	Any	Required (1 or more)	Yes	Yes	None
Combo	Any	Optional	Yes	Yes	None
Histogram	All Time only	Optional (max 1)	No	Yes	Binning required
Pie	All Time only	Required (1 or more)	Yes	Yes	None
Treemap	All Time only	Required (1 or more)	No	No	Formula for compound measures
Waterfall	All Time only	Required (1 or more)	No	No	Formula for compound measures
Scatter	All Time only	Optional	No	No	3 measures (X, Y, size)
Heatmap	All Time only	Required (exactly 2)	No	No	Formula for compound measures
KPI	Any	Optional	Yes	Yes	None
Impact	Any	Optional	No	Yes	Trigger event required

MCP prompt examples

“Create a stacked bar chart of daily tool calls by primary intent for the last 7 days”

“Show me a heatmap of average friction score by primary intent and friction signal for the last 7 days”

“What is the impact on Place Order count after a user triggers Add to Cart, last 7 days?”

Query report: build measures and prompt Kubit through MCP.
Formula: combine measures for chart types that need a single value.
Chart settings and formatting: axes, formats and reference lines.

Start here

Agent Observability

Product Analytics

Bring your own warehouse

Query chart types

When to use

Common parameters

Time Series charts

Bar