Use AI Studio to explore customer data before you build a campaign. Marve can inspect schemas, validate field values, run queries, summarize patterns, and return charts when a visual comparison helps.

When to explore data

Use data exploration when you need to answer questions like:

• Which fields can help identify churn risk?
• How many customers purchased in the last 90 days?
• Which product categories are growing fastest?
• Which cohorts have lower engagement than expected?
• What values does this status or category field actually contain?
• Which existing audiences are connected to a destination?

Ask a focused question

Start with the business decision you need to make. Include the population, metric, and timeframe when you know them.

Review the answer

Marve can use warehouse queries, GrowthLoop metadata, and charts to answer your question. Review the result before you use it in a campaign plan.

Check:

• The selected dataset group.
• The fields and joins Marve used.
• The date range.
• The row count or sample size.
• Any caveats about missing fields, null values, or incomplete data.

📘

Note

Ask Marve to show the fields, filters, or query logic it used when you need to validate an answer with your data team.

Move from insight to action

Use the answer to decide what to build next.

Explore data
  |
  v
Find a cohort, trend, or field
  |
  v
Ask Marve to generate an audience
  |
  v
Review audience logic in GrowthLoop

Example follow-up prompts:

• "Turn that cohort into an audience I can review."
• "Create an audience for the highest-risk group from this analysis."
• "Add a 90-day purchase window and exclude customers with open renewal opportunities."
• "Chart that trend by month so I can compare the cohorts."

Next, learn how to generate audiences from a prompt or data exploration result.