Skip to content
English
  • There are no suggestions because the search field is empty.

Best Practice Guide: How to ask questions in ASK BOSCO® AI Analyst.

ASK BOSCO® AI Analyst is a powerful tool for finding insights from your data. Like any good analyst, it performs best when you provide clear, thoughtful instructions. By learning how to frame your questions effectively, you can get to the answers you need faster and with greater accuracy.

Navigation

New ASK BOSCO® AI Analyst  - March 2026

Updated Best Practice

Troubleshooting

Further Reading


What's new in the March 2026 Update to ASK BOSCO® AI Analyst?

AI Analyst has introduced significant improvements over previous versions:

  • Autonomous analysis — Translates natural language to data queries, plans analysis, and delivers insights.

  • Narrative insights: Provides rich text summaries explaining why your data looks the way it does. Provides insights, not just charts.

  • Automatic Model selection: Identifies and queries across multiple data models at the same time and selects the most relevant data source(s) based on your question.

  • Clarification requests — Asks questions when uncertain instead of guessing.

  • Data-awareness: Now securely reads the specific results of your query to write accurate insights and answer follow-up questions with full context.

  • Insight storage and chat history: Generates insights from your data and stores them as your conversation history within AI Analyst.

  • Data security: Processes only the data needed to answer your question and the system operates strictly within a defined dataset and analytical framework.

    The AI cannot act outside its configured data environment, and all outputs are directly traceable to the underlying dataset. Your data is used only to generate insights and is never stored or used for training. 

  • Analyst-grade, transparent reasoning: Creates a plan, validates assumptions, and self-corrects to provide you verifiable results. Shows analysis steps in plain language.

  • Multi-turn conversation: You can ask follow-up questions, request explanations, and explore the "why" behind the numbers. Its smarter context detects new questions, versus follow-ups, automatically.

AI Analyst March 2026
Back to top


Our best practices to help you master your conversations with AI Analyst.

Ask effective questions

Be specific about your intent.

AI Analyst handles open-ended questions but uses its own judgment to plan analysis. This could result in you not getting the answer you were aiming for. Specify dimensions explicitly if needed.

Less effective

More effective

"How are sales doing?"

"How are sales doing? I expect analysis at channel and campaign level."

"Show me device data."

"Show me device traffic and cost by channel."

"What’s our performance?"

"Revenue performance this quarter. Compare it against last quarter and same quarter last year."

Use Research Mode for Complex Analysis

AI Analyst Research Mode




Research mode introduces a process where AI Analyst reasons through a problem like a human analyst.

  • Extended planning: Unlike the default Search mode, Research Mode creates a multi-step query plan to simplify complex questions. It identifies ambiguities, generates testable hypotheses, and allows users to modify the analytical path before execution.

  • Iterative self-correction: If an anomaly is detected, the agent autonomously deep-dives, self-corrects its logic, and refines subqueries until it reaches a grounded conclusion.

  • Automated narrative summary: The final output is a coherent report that explains the “why” and “what’s next,” citing specific data points and visualizations to build trust, structured in a detailed report.

 

Search Mode

Research Mode

Reasoning/Thinking time

Low: Optimized to be fast

High: Optimized to be comprehensive

Prompt depth

Single-dimensional (“What”)

Multidimensional (“Why” and “what if”)

Output type

Direct answer or visual

Comprehensive decision report

Use Auto Mode

 Use Auto mode if you don’t know which Data Model to use. 

AI Analyst Auto Mode

Auto mode also supports cross-source queries in case you want to query data from multiple sources to generate insights

Manual data model selection improves accuracy when models have overlapping data. Select the model, not Auto mode, if you know which model your question relates to.

Brainstorm with AI Analyst

Provide role context to get analysis suggestions:

"I'm a Paid Social account manager focused on awareness driving. What metrics should I review?"

"I am a Buying and Merchandising Director. What data will help me to maximize margin?"

"I manage Affiliates. What questions should I ask about our partners performance?"

Back to top

Troubleshooting

ASK BOSCO® AI Analyst is a powerful tool, but its outputs should always be evaluated critically.

Unsure about a response

  1. Ask AI Analyst to explain its analysis.
  2. Ask AI Analyst to verify its analysis or double-check a specific result.

Request clarification

If results seem incorrect, ask AI Analyst to explain:

"How did you calculate that percentage?"

"Why did you exclude those channels?"

"Double-check this calculation"

Identify non-data responses

AI Analyst can answer open-ended questions, but ambiguous queries may produce responses from general knowledge instead of your data.

Signs of a non-data response:

  • Generic-sounding answer
  • No specific numbers from your data
  • Reasoning block shows no data queries

When in Deep Analysis

Always review the analysis plan before approval. Add specific dimensions or comparisons you need.

  • Expect longer processing time for thorough analysis.
  • Remove irrelevant steps to save time and cost.
  • Don’t close the tab while running Research mode to ensure AI Analyst finishes its analysis.

Auto mode

Works best when the overlap between available data models is low. It may return options for which models it can use to answer your question, and you can choose.

  • For overlapping models, specify the data model name if you already know it.
  • Ask AI Analyst to run queries on multiple models if you cannot choose.
  • Ask AI Analyst to share more details about suggested data models if you need help deciding.

Back to top


Further reading on questions in ASK BOSCO® AI Analyst can be found here

Article - A Guide to asking questions in ASK BOSCO® AI Analyst

Back to top


That completes our Best Practice Advice. If you have any questions please contact team@askbosco.io