Chatting & getting good answers

Chatting with a Norami assistant is like messaging a colleague who knows your data inside out. You ask in plain language; it answers from your actual numbers.

Asking a question

Open an assistant and type your question. You can ask for facts, summaries, comparisons, or breakdowns:

How many orders did we have last month?

Compare revenue by region for this year.

Which 10 clients spent the most, and how much?

The assistant reads your data, works out the answer, and replies — often with a table or a short explanation of how it got there.

Follow-up questions work

The assistant remembers the conversation, so you can drill in naturally:

...now show only the south region.

What about the same period last year?

Tips for great answers

• Be specific about the time frame and grouping. "by month", "in 2025", "per salesperson" all help.
• Name the metric you mean. If your data has several amount columns, say which one ("net revenue", not just "amount").
• Ask for the breakdown if you want it. "show the numbers" or "as a table" gets you the detail behind a summary.
• One question at a time tends to beat one long multi-part question.

If an answer looks off

• Rephrase with more specifics — the time range, the exact column, the grouping.
• Check that the underlying data is current (see Updating & refreshing data).
• Remember the assistant answers from the data it has — if something isn't in your uploads, it can't know it.

Going further