Recollective AI: How to Ask Great Questions

The Recollective platform includes an AI-powered feature called AI Questions, allowing you to ask questions about your collected research data and get instant answers and insights.

The AI Questions feature has three main components:

• Question: You write a question or prompt based on the information you want to learn from your data.
• Answer: The AI analyzes your research data and generates an answer to your question.
• Verbatims: The AI includes direct quotes from your participant responses ("verbatims") that support the AI's answer. This allows you to assess the quality of the answer and find supporting quotes for your reports.

Strategies for Asking Effective Questions

For ideas on how you can leverage AI Questions, take a look at the slides below! 

Click the three dots to view in full-screen.

In addition to the examples above, the following strategies will help you get the most out of the AI Questions feature. 

1) Define the Scope of Your Question

• Focus on Specific Data: Use the filters or @ mentioning to focus the AI on specific activities, tasks, participants, or segments for more accurate answers.
• Keep Questions Relevant: The AI can only respond based on the data from your study participants, so make sure your questions are relevant to the study's scope.

2) Craft Clear and Specific Questions

• Use Open-Ended Questions: Ask open-ended "Why" or "How" questions for deeper insights, avoiding quantitative or yes/no questions.
• Provide Context: Be clear and detailed in your questions, specifying terms to avoid ambiguity and help the AI understand your intent.
• Be Specific About Brands and Media: Clarify brand or media references by adding context, and use quotation marks for names to guide the AI. (Ex. The movie “Titanic”)
• Specify Your Desired Output: Include instructions for how you want the AI to format the response, such as bullet points, lists, markdown tables, or summaries.

3) Experiment and Iterate

• Start Broad, Then Narrow Down: Begin by asking broad questions to get general insights. Then, based on the AI's initial answers, ask more specific follow-up questions to uncover deeper insights.
• Try Different Phrasings: Experiment with different ways of phrasing the same question to help uncover additional insights and perspectives. The AI might provide different answers or identify different verbatim quotes based on how the question is phrased.
• Explore Different Question Types: Experiment with asking for opinions, experiences, or predictions to see which questions generate the most helpful insights.
• Don't Be Afraid to Ask Follow-Up Questions: Ask follow-up questions based on the AI's responses to clarify any ambiguous points or dig deeper into interesting findings.

Additional Tips for Effective Questioning

• Focus on What's There, Not What's Missing: Ask questions about information that is included in your data, not what is excluded.
• Avoid Emojis and Random Phrases: The AI is optimized to handle questions related to your study data. We recommend avoiding characters or phrases that may produce unexpected results. 
• Use @ Mentioning to Make Comparisons: Mention multiple segments or participants to compare insights against each other. 
• Let the AI Guide You: If you're not sure where to start, you can ask the AI to suggest questions based on your study data.
• Use Multiple Chats: For better results, start a new chat when changing topics. This helps the AI focus on the specific subject without confusion.

Providing Feedback

The Recollective team is dedicated to continually improving the AI Questions feature. You can help by providing feedback directly within the platform. After asking a question, use the thumbs-up or thumbs-down icons to rate the helpfulness of the AI's answer. You can also leave a comment to provide more detailed feedback.