Asking Questions About Your Data
The Database Assistant allows you to ask questions about your data using natural language. It translates your questions into SQL queries and returns answers based on actual database results. This article shows how to ask effective questions and includes example queries you can reuse.
Written By Kristė Vagnerytė
Last updated 23 days ago
Start with a Clear Business Question
Begin by describing what you want to understand about your data, not how to query it.
Good examples:
“What are our sales by month?”
“How many active users do we have per day?”
“Which products generate the highest revenue?”
You do not need to write SQL — the assistant generates and executes queries automatically.
Example: Exploring Trends Over Time
Use questions like this when you want to understand performance changes over time.
Example request:
What are our sales by month? Create a line chart.
This helps identify trends, seasonality, or unusual changes in performance.
Provide Context to Improve Accuracy
Adding context makes results more precise. When possible, include:
The metric (sales, revenue, users)
The time period (last month, this year, specific dates)
The dimension to group by (month, region, product)
Example request:
What are our total sales by month for the past year?
Example: Understanding Growth Drivers
Use this type of question when you want to understand why results changed.
Example request:
What were the main growth drivers in 2025?
This helps analyze which factors contributed most to changes in performance.
Example: Customer Segmentation
Use segmentation questions to understand how different groups contribute to results.
Example request:
What portion of sales comes from new clients versus repeat clients?
This helps analyze customer mix and behavior.
Requesting Charts or Summaries
By default, the Database Assistant returns answers as text.
If you want a chart or summary, ask explicitly.
Example request:
What are our sales by month? Create a line chart showing the trend.
Charts and documents are created only when you request them.
When Results Are Not What You Expected
If the result seems incomplete or unclear:
Clarify the metric or timeframe
Ask the assistant to explain how the result was calculated
Narrow the scope of the question
Small adjustments usually lead to better results.