AI customer assistant
Amazon Analytics is a conversational interface for analyzing buyer and order data. Ask questions in plain English and get answers without writing SQL. It’s not a chatbot for your customers (that’s Hilal Chatbot, a separate product) — it’s a tool for you to interrogate your buyer dataset.
In this guide:
- Access Amazon Analytics
- Ask a question about your buyers
- Copy and save responses
- Understand what data is available
Step 1: Access Amazon Analytics
Go to Customer Hub → AI Assistant in the sidebar. The page displays as “Amazon Analytics” at the top.
/customer-hub/ai-assistant/db. It is not linked anywhere else in the navigation today.Step 2: Ask a question
Type a question about your buyer and order data. Amazon Analytics translates your question into a query and shows results. Examples:
- “How many repeat buyers do I have?”
- “What’s my refund rate by SKU?”
- “Show me buyers who’ve spent more than $500 lifetime.”
- “Which marketplace has my highest review rating?”
Results display as numbers, tables, or summaries depending on your question.
Step 3: Copy responses
For each answer, you can copy the underlying SQL query to your clipboard to use in other tools or for reference.
Step 4: Understand the data
Amazon Analytics queries your NL2SQL database backend to answer questions about your business data.
What data is available
Amazon Analytics can query data available in your NL2SQL database:
- Orders and order line items.
- Refunds and returns.
- Reviews and ratings.
- Buyer messages.
- Marketplace and product dimensions.
customer_reply_ai feature entitlement. The scope of available data depends on your backend database configuration.Limitations
Troubleshooting
- “I can’t answer that.” Your question may be outside the available data scope, or Amazon Analytics didn’t understand. Try rephrasing more concretely (“how many” / “what percentage” / “show me a list”).
- Answer seems wrong. Try asking it to show the underlying query. If the query is correct but the data seems unexpected, there may be a data quality issue.