Customer Knowledge Base

Chatting with Emma (How to Use)

πŸ§ͺ Emma is currently in beta
The AI Assistant is available to selected test users only. If you do not see the assistant icon in your navigation bar, your account has not yet been enabled. Please contact your emax digital account manager to request access.

Emma responds to free-text questions, can read data from the Product Performance board currently visible on screen, and formats its answers to make analysis fast and actionable. This article explains how to send messages, how context awareness works, what a typical response looks like, and what you can do with each response.

Sending a Message

The message input field is located at the bottom of the Emma panel. It displays the placeholder text "Ask anything..." and accepts free-text questions, instructions, and Markdown.

To send a message:

  1. Click the input field and type your question or instruction.

  2. Press Enter or click the ↑ Send button (the blue arrow icon, which highlights once text is present).

04_emma_message_input-20260610-115344.png

To the left of the input field is the ⊞ Browse Skills button. Clicking it β€” or typing / in the input field β€” opens a skill shortcut picker. See the Skills (Shortcuts) article for details on using skills.

Context Awareness β€” Using On-Screen Data

Emma has context awareness of the Product Performance board β€” the ASIN table, charts, and filters currently visible on screen, as configured by the user. Emma reads this data automatically without any additional setup. You do not need to copy or paste any data manually.

To reference the on-screen data, simply describe what Emma should analyse in your message. Emma will interpret the visible content and respond accordingly.

Context awareness is scoped to the Product Performance board. Emma reads whatever is displayed there β€” including Sales, Inventory, Advertising, Content Score, Buy Box, and Ratings data β€” depending on how the user has configured the view.

Example prompt:

Use the displayed product performance table on the screen and give me a recommendation on what to do next.

How a Conversation Works

The diagram below shows the three ways to initiate a message, how Emma responds, and the actions available on each response.

Diagram showing the three ways to start a conversation with Emma, her response format, and the action buttons available

Response Format

Emma formats its responses for quick reading and analysis. Responses can include:

  • Markdown styling β€” headers, bold and italic text, bullet points, and numbered lists

  • Emojis for visual labelling (e.g. πŸ“Š for performance data, βœ… for positive insights, ⚠️ for warnings, 🌍 for marketplace breakdowns)

  • Tables for structured data such as marketplace performance or advertising metrics

  • Coloured status indicators β€” 🟒 green for positive signals, ⚠️ orange for areas requiring attention

At the end of each response, Emma proactively suggests follow-up actions.

Response Action Buttons

Below each response, a set of action buttons is displayed alongside the response timestamp.

Button

Function

πŸ‘ Thumbs Up

Rate the response positively

πŸ‘Ž Thumbs Down

Rate the response negatively

β“˜ Info

Displays token usage: Input tokens, Output tokens, and Total tokens

πŸ“‹ Copy

Copies the response text to clipboard

Copy Markdown

Copies the response in raw Markdown format

⬇️ Download

Downloads the response content

πŸ“„ Download PDF

Downloads the full response as a formatted PDF document

πŸ“Š Download Excel

Exports individual tables as Excel files. Only available when the response contains structured table data.

Note: AI-generated content can be inaccurate. Always verify critical information before making business decisions.