Learn

Explore your data in Mammoth: which tool to use

Mammoth has four tools that show what is in a View before you build any transformation: Explore Cards, the Metrics panel, View Insights and the Data Quality Report. You find them all in the data editor: the Explore toggle and Metrics icon in the action bar, the View Insights button on the right of the action bar, and View Data Quality in the footer below the grid. Viewers can read Explore Cards; filtering needs edit access to the workspace. View Insights and the Data Quality Report are unavailable in preview mode, in edit mode and in a view-only workspace.

Animation of a manufacturing data View with Explore Cards for shift, scrap_count, temperature_celsius and date above the grid

Which page do I need?

I want to...	Go to
See what is in a column and filter the grid by clicking values	Explore Cards
Filter by a sentence, search every column, or plot two number columns	Ask, Search and Compare cards
Read a text, number or date card, change its grouping or measure	Text, number and date cards
Search, hide, show or reorder columns, or rename and retype them	Browse columns
Track a total, average or custom KPI while I filter	Metrics
Get an AI summary of what is interesting in my data	View Insights
Check data quality and get suggested fixes	Data quality report
Read, search and filter a View on my phone	Use Mammoth on a phone

The four explore tools at a glance

View interface with the Explore toggle and Explore Cards panel, the 5 Key Insights button, the Metrics panel with a Sum of quantity metric, and the Data Quality Score footer

Tool	What it gives you	How you interact	What you can do with the result
Explore Cards	Text: a ranked list (or donut, treemap). Number and date: a vertical bar chart (or list; timeline for dates)	High: click values to filter	Turn filters into a pipeline task or a new View; add a card's chart to a dashboard
Metrics panel	Single numbers, such as a sum or a custom formula	High: formulas and conditions	KPI calculations and dashboard preparation
View Insights	3–7 AI-written observations	None: read only	Informational only
Data Quality Report	A score, six dimension scores and recommended tasks	Low: read only	You build the recommended tasks in the pipeline yourself

Comparison of the four discovery tools by analysis time, output type, interactivity and actionable result

Pick an explore tool by your goal

Goal	Start with
I loaded new data and want to understand it	Explore Cards
I need to check data quality before building a pipeline	Data Quality Report
I want a quick summary of key numbers	Metrics panel
I want AI to tell me what's interesting in my data	View Insights
I want to investigate a specific pattern or segment	Explore Cards, then filter by clicking values

Decision tree: Quick Look, Find Patterns, Check Quality and Calculate KPIs lead to four different tools

Use Explore Cards for a first look and filtering

Explore Cards show a card for each column you add. Text columns default to a ranked list of distinct values with counts and percentages, and can switch to a donut or treemap. Number columns default to a vertical bar chart of value ranges. Date columns default to a vertical bar chart and can switch to a timeline. Number and date cards can also show a list.

Good for: a first look at new data, value distribution checks, spotting outliers, and testing filter logic before you automate it.
Not for: AI business observations, a full quality score, or calculations beyond the card statistics (Sum, Min, St.Dev, Avg and Max for numbers; Earliest and Latest for dates).
Example: "I uploaded sales data and want to see the spread by geography before I build regional reports."

Explore panel showing a text card, a numeric card and a date card

Double-click a value to filter at once. The filter badge beside the Explore toggle counts your active filters, and you can later turn them into a permanent pipeline task.

Use the Metrics panel for specific numbers

The Metrics panel creates single-value metrics and KPIs. They update with the View and your explore filters, unless you lock a metric to keep its value fixed. It prepares numbers for dashboards but does not create dashboards.

Good for: specific KPIs, ratios and percentages, watching a number as you filter, and comparing aggregations.
Not for: basic column statistics (Explore Cards already show them) or full data profiling.
Example: "I'm building a sales dashboard and need conversion rate, average deal size and quota attainment."

Metrics panel with three metrics: Sum of Price, Sum of Tax filtered by Department, and Minimum Quantity

Use View Insights for an AI summary

View Insights reads the current View and writes 3–7 business observations about patterns, trends and anomalies. The AI picks the count, and you can't set it. Before the first run, the button on the right of the action bar reads View Insights; afterwards it shows the count, for example 5 Key Insights.

Good for: briefings, finding trends you didn't know to look for, and "what should I know about this data?"
Not for: quality scores or interactive exploration.
Limits: no follow-up questions, no export button (copy the text yourself), and the current View only.
Example: "I need to present quarterly results to the board and want the main trends highlighted."

View Insights panel under the 7 Key Insights button, listing seven insights about customer revenue and churn

Use the Data Quality Report before you rely on the data

The Data Quality Report scores the current View on six DAMA (Data Management Association) dimensions and recommends pipeline tasks, such as Remove Duplicate Rows, Conditional Filter, Text Formatting or Math Function, labelled High Priority, Medium or Low. Open it from View Data Quality in the footer below the grid; View Details opens the full report. There is no one-click apply: you build the tasks yourself.

Good for: checks before production, audit and compliance evidence, and planning which cleaning steps to build.
Not for: quick exploration, or issues you already know and accept.
Limits: no export (screenshot or copy), a snapshot rather than continuous monitoring, the current View only, and no custom dimensions beyond the DAMA six.
Example: "This data feeds our finance dashboard. I need to check it before executives see it."

Data Quality dialog Recommended Actions section listing Edit Cells and three Fill Missing Values cards with High Priority and Medium labels

Run the report as soon as you connect a new data source, and fix critical issues before you spend time on a pipeline.

What each explore tool sees and when it refreshes

Tool	Data it reads	When it refreshes
Explore Cards	The View after its pipeline, narrowed by your explore filters	As you filter and as the View's data changes
Metrics panel	The same, unless a metric is locked	As you filter and as the View's data changes
View Insights	The whole View after its pipeline; explore filters don't change it	Only when you click the button. A saved result is cleared when the View's data changes
Data Quality Report	The whole View after its pipeline; explore filters don't change it	Only when you click the button. A saved result is cleared when the View's data changes

To make View Insights or the Data Quality Report look at a subset, add a Conditional Filter task to the pipeline. See Filter rows.

Combine the explore tools in a workflow

New dataset:

Explore Cards to see structure and distributions.
Data Quality Report to check the data is ready.
View Insights to find business patterns.
Metrics panel for the KPIs you need.
Turn what you found into pipeline tasks.

Briefing for a manager or board:

View Insights for the main trends.
Metrics panel for the numbers behind each point.
Explore Cards to confirm the patterns visually.
Build a dashboard from the metrics.

Production pipeline:

Data Quality Report to find what needs fixing.
Explore Cards to understand the columns you will transform.
Build the pipeline tasks.
Metrics panel (optional) for success measures.
Run the tools again to confirm the improvement.

Examples by industry

Financial services, monthly regulatory reporting: run the Data Quality Report for completeness and accuracy, check unusual transactions with Explore Cards, calculate key ratios in Metrics, then use View Insights for a summary for the compliance team.
Manufacturing, production line analysis: check defect rates by product line with Explore Cards, let View Insights point out patterns, calculate yield in Metrics, then confirm sensor data with the Data Quality Report.
Hospitality, revenue management: use View Insights for seasonal patterns, filter peak and off-peak periods with Explore Cards, calculate occupancy, RevPAR and ADR in Metrics, then check booking data completeness with the Data Quality Report.

Choose an explore tool for common situations

Four scenario flows with ordered tool sequences, from 10-Minute Analysis to Unknown Questions

Situation	Order to use the tools
"I have 10 minutes to understand this dataset"	Explore Cards, View Insights, a few metrics, then the Data Quality Report. Use any time left to plan next steps
"My boss needs insights for tomorrow's meeting"	View Insights first, Metrics for the numbers, Explore Cards if questions come up. Skip the Data Quality Report unless quality is on the agenda
"This data feeds our production dashboard"	Data Quality Report, Explore Cards for thresholds, Metrics for exact KPIs. View Insights is optional
"I'm not sure what questions to ask"	View Insights first, Explore Cards to follow its leads, the Data Quality Report for limits, then Metrics for ratios you find

To check a finding, use more than one tool. For example, View Insights says "Revenue grew 23% in Q3". Filter to Q3 in Explore Cards, calculate month-over-month growth in Metrics, then check revenue completeness in the Data Quality Report.

Turn what you find into action

Explore Card filters become a Conditional Filter task or a new View. See Explore Cards.
An Explore Card's chart goes on a dashboard from its More options menu (Add to dashboard…).
Data Quality Report recommendations become tasks you build in the pipeline.
View Insights patterns are leads to research with Explore Cards. Copy the text for documentation or talking points.
Metrics feed dashboards.
Data Quality Report scores work as quality gates in your pipeline process.
View Insights output can document a dataset, and findings can be copied or screenshotted for stakeholders.

Troubleshooting explore tool choices

Symptom	Cause	Fix
"I used View Insights but need specific numbers"	View Insights gives observations, not calculations	Use the Metrics panel
"The button showed 5 Key Insights but now I get 3"	The count is set by the AI each time, between 3 and 7, and can differ on the same data	Nothing to fix. You can't set the count
"The Data Quality Report found issues but I need quick insights"	Patterns on dirty data can mislead	Build the recommended tasks for critical issues first, then run View Insights
"Explore Cards don't show the patterns I expected"	A card to the right of a filtered card shows only the filtered rows (its header notes this), or the column has no card	Use Clear filters, and check you added the columns you want
"The tools give different numbers"	Explore Cards and Metrics follow your explore filters; View Insights and the Data Quality Report don't	Clear the explore filters, or add a Conditional Filter task so every tool reads the same rows
"I don't have time to run all four"		Quick overview: Explore Cards. Briefing: View Insights and Metrics. Production check: Data Quality Report. Full analysis: all four in sequence

How the pipeline works: where your explore findings become repeatable tasks.
Filter rows: the Conditional Filter task that explore filters turn into.
Create a dashboard: show your metrics and charts to others.

Was this page helpful?

Something wrong? support@mammoth.io