Overview

The Insights Agent is an AI-powered analytics assistant that helps you understand how your search, browse, autosuggest, and merchandising campaigns are performing, using simple natural language questions.

Instead of navigating multiple reports or relying on analytics teams, users can ask questions directly and get instant insights.

It is available within the Reporting section, accessible across:

Search reports
Browse reports
Autosuggest reports
Campaign and merchandising performance reports

The agent appears as a widget within reporting pages and opens a chat interface when activated.

Working of Insight Agent

The Insights Agent analyses existing reporting data in the console and presents it in a summarised, easy-to-understand format. Instead of requiring users to apply filters, switch between reports, or manually derive insights, the agent interprets natural-language questions and automatically retrieves the relevant data.

If a question is unclear, such as whether it refers to search or browse performance, the agent asks follow-up questions to clarify the context before generating a response. Users can continue the conversation with follow-up questions to drill deeper into the data.

Use Case

Refer to the use case below to understand in which scenario what support does Insight Agents provide.

Use Case	User Scenario	What the Insights Agent Does	Outcome
Search Performance Analysis	A user wants to understand how search contributed to business performance over a specific time period.	The agent analyses search data and summarises key metrics such as sessions, clicks, carts, orders, and revenue. It can also compare performance across different periods, like quarter-over-quarter or year-over-year.	The user quickly understands overall search performance and trends without having to navigate multiple reports.
Query Performance & Search Quality	A user wants to identify which queries are performing well and which ones need optimisation.	The agent evaluates query-level data to identify top queries by revenue or conversion, highlight high-traffic but low-conversion queries, and analyse head, torso, and tail query behaviour.	The user gains visibility into search quality and can prioritise optimisation efforts for high-impact queries.
Merchandising & Campaign Impact Analysis	A user wants to measure whether a promotion or campaign improved performance.	The agent compares performance metrics before and after a campaign. It evaluates campaign-level impact and highlights which promotions are driving results.	The user can assess the effectiveness of merchandising efforts and make data-driven decisions for future campaigns.