Introducing Marvin deep research: Get automated insights

You asked for quicker, deeper insights, and we listened!

We’re excited to officially introduce deep research, formerly known as auto analysis, your AI-powered deep dive into insights.

What is deep research?

Deep research is your AI-driven analysis tool, designed for effortless and comprehensive research analysis. A “set it and forget it” feature that you can initiate to automatically process your entire project’s data and give you custom structured reports with minimal effort.

Choose the type of Deep Research analysis you want to run and provide your prompt/topic/keyword/question. Then, Marvin AI performs an in-depth analysis, reasoning through all your research files in a project, intelligently interpreting and adapting to the information it finds. Think of it as handing over your raw research data to a highly intelligent research assistant who returns a detailed, insightful report in minutes. Deep research can process, analyze, and generate reports in tens of minutes, saving hours of research bandwidth.

Why “deep research”? A name that reflects its true power

You might be familiar with this feature as “Auto analysis.” We renamed it to “Deep research” because this new name more accurately captures its core strength: delivering truly deep and meaningful insights.

Though “Auto analysis” was a functional name, it didn’t showcase the depth of AI exploration and sophistication that this feature provides.

The name “Deep research” helps to reflect its ability to:

• Dive deeper into your data: Look beyond surface-level findings and uncover hidden patterns and nuances.
  
• Conduct comprehensive analysis: Intelligently analyze all files in your project for a holistic understanding.
  
• Empower in-depth understanding: Provide you with the deep insights you need to make informed decisions.

Why use deep research?

Deep research offers an unmatched, powerful combination of speed, depth, and ease of use.

• Hands-free research reports: Go from data to insights with minimal effort from the researcher. Simply upload your data into your project and run deep research to create structured reports summarizing your research automatically.
  
• Quicker insights: Accelerate product decisions. Get detailed reports and key insights in minutes and streamline your research process.
  
• Thorough data coverage: Deep research is designed to sift through each and every file in your project(up to 75 files), ensuring no insight is missed.
  
• Automated theme discovery: Automatically identify recurring themes and topics, even subtle ones, across your entire dataset.
  
• Custom structured reports: Receive detailed reports, from quotes to overarching themes, with supporting evidence based on your analysis type.
  
• Handles raw data: Works directly with raw data (transcripts, surveys, etc.), eliminating the need for manual pre-processing, tagging, or note-taking.
  
• Time and resource savings: Free up valuable researcher time by automating data processing and report generation.

What can you do with deep research?

Deep research is equipped to perform a range of powerful analysis tasks:

• Keyword-driven exploration: Effortlessly uncover relevant information by simply entering keywords related to your research questions.
  
• Rigorous hypothesis testing: Formulate hypotheses and use deep research to test them rigorously against your project data.
  
• Pinpoint essential quotations: Quickly locate impactful quotes relevant to your research topics and extract key statements that support your analysis.
  
• Data-driven thematic analysis: Conduct quantified thematic analysis to identify and measure the prevalence of themes within your dataset.
  
• Develop rich user personas: Generate detailed personas based on your data to better understand user segments and behaviors.
  
• Structured Q&A for clear insights: Create structured question-answer tables to get a summary of how each participant answered your questions.
  
• Create custom reports: Need to explore a specific angle? Utilize custom prompts to ask tailored questions and investigate exactly what you need to know.
  
• Make richer insight reports: You can add your findings within deep research into Insights and integrate them into your broader research workflow.
  
• Filter, compare & discover : Filter and narrow down the files used in your research to focus on specific data segments. You can also compare them to discover common themes or pinpoint key distinctions across segments.

How does deep research work?

Deep research leverages advanced AI to analyze your data in a streamlined, automated process:

1. Data ingestion: You upload your research data (transcripts, surveys, etc.) into your Marvin project.
   
2. Specify analysis: You initiate deep research and select the type of analysis you want to run (e.g., “Uncovered Topics,” “Quotes,” “Insight Report”).
   
3. AI-powered processing: Deep research’s AI engine automatically processes all files in your project, identifying relevant information based on your chosen analysis type, the keyword/prompt you enter and underlying algorithms.
   
4. Insight extraction: The AI extracts key themes, relevant quotes, and other insights based on the analysis parameters.
   
5. Report generation: Deep research compiles these insights into a structured report with supporting evidence (Data Source).
   
6. Results delivery: You receive a detailed, AI-generated report ready for review and further action. 

Deep research vs. Ask AI: Know your AI assistants

While both deep research and Ask AI are part of your AI-powered research repository, they serve different purposes:

Deep research vs. Analyze page: Choosing your approach

Deep Research and the Marvin analyze page both offer powerful research analysis, but with distinct approaches:

Use cases for deep research: Real-world applications

Deep research is incredibly versatile. Here are just a few examples of how you can use it:

• Customer feedback analysis: Quickly analyze large volumes of customer feedback (surveys, transcripts) to identify recurring pain points, feature requests, and sentiment trends.
  
• Market research reports: Generate initial drafts of market research reports, identifying key market trends, competitor insights, and customer needs.
  
• Competitive analysis: Analyze competitor data to understand their strengths, weaknesses, and market positioning.
  
• Persona development: Use deep research to analyze user interviews and identify common characteristics to kickstart persona creation.
  
• Project overviews for stakeholders: Generate concise reports summarizing key findings from a research project for quick stakeholder updates.
  
• Analyzing historical data: Quickly extract insights from archives of past research projects to identify long-term trends or recurring issues.

Deep research best practices: Do’s & don’ts

Designed to deliver automated insights, deep research works best when you guide it by choosing the right template for the right prompt. We must remember that AI can never be a replacement for a researcher. To help AI work with your research data best:

• Add segmentation data using file tags: Add file tags to guide AI in understanding which segment the query concerns and use specific prompts to receive accurate and comprehensive reports.
  
• Experiment with different analysis types: Explore “Uncovered Topics,” “Quotes,” and “Q&A” to see which best suits your needs.
  
• Use custom prompts for specific questions: Guide deep research with custom prompts to explore specific areas of interest.
  
• Review and refine reports: Deep research provides a fantastic starting point, but always review and refine the AI-generated reports with your own expertise and insights.
  
• Leverage deep research for efficiency: Embrace its “set it and forget it” nature to save time on initial data analysis and report generation.
  
• Utilize data sources: Cross-check the AI-generated report against the data source tab and dive deeper into the details from each file.

Effective Deep Research requires a balance: knowing both how to ask and what to expect from its AI-powered analysis. Don’t:

• Expect perfect quantification of themes: Deep research excels at theme identification, but nuanced quantification may require further manual analysis.
  
• Rely solely on deep research for highly customized analysis: For a more deeply interactive and researcher-driven analysis, the Analyze page is more appropriate.
  
• Forget to add your expertise: AI is a powerful tool, but your researcher expertise is crucial for interpreting and contextualizing the insights generated by deep research.

Unlock deeper insights, effortlessly

Deep research is your AI-powered shortcut to profound, structured UX insights. Whether you’re a trained researcher or a non-researcher, the feature allows you to leverage its automated power to generate full research reports hassle-free. You can uncover themes, create custom reports, and better understand your data. Deep research takes over research analysis to free you up to focus on strategic thinking and impactful decision-making.

Start exploring deep research today and experience the future of efficient, in-depth analysis!
Book a free demo to see Marvin’s deep research in action.