How to conduct cross-project analysis

With cross-project analysis, you can view, compare, and analyze notes across all projects in one place.

Run AI analysis and add notes to Insights, Playlists, and more.

Select More research tools in the left navigation menu and click Cross-project analysis.

Here you'll see notes from all projects on this page. Similar to the Analyze notes page, you can filter notes by file tags, labels, participant data, creation date, project, etc. Select any filter in the toolbar at the top of the page to apply filters.

Select the notes you want to analyze together. Marvin will run AI analysis only on the notes you select. If you don't select any, it will run analysis on all notes.

If you have thousands of notes across projects, we recommend that you add filters and search among a smaller dataset to get more accurate answers.

Choose the type of analysis based on the answer you're looking for.

Test a hypothesis or ask any question about the selected notes. Just type your question in the toolbar at the top of the page. This is similar to running Ask AI elsewhere in Marvin. Marvin will consider the transcripts and the notes to give you an answer.

This can be particularly useful when you want to uncover topics that repeat across multiple files.You may choose to use it to analyze usability testing calls, feedback calls, demos, etc. For example, we ran this analysis for the notes we’ve made on all Marvin webinar recordings.

Marvin’s AI was able to detect the themes that one or more of our guest speakers mentioned. It searched all the selected notes and grouped relevant information under each theme it found.

Quick thematic analysis can be a great starting point for your research. All the themes and sub-themes that Marvin identifies may become points where you want to delve deeper. You may then choose to use the Advanced thematic analysis.

Marvin will highlight how many times a theme was mentioned in your selected notes. You can open the analysis in a new tab if you want to keep referring to it. You’ll also see options here to add the analysis to an Insight report, copy the text or regenerate the analysis if you’re not happy with it.

We recommend that you check all AI generated results before presenting your research to stakeholders

The advanced version of thematic analysis goes a step further, identifying an exhaustive list of topics and adding more detail to each of them. It also lets you explore the sources used to identify each theme.

When you click “Mentioned in 12 notes” under “User-centric Approach”, you can read all the corresponding notes. Marvin also creates a graph for you to visualize the data and present it in your reports.

You may also want to stitch together multiple notes into a Playlist to make the point more comprehensive.

You can multi-select notes and add labels too. This helps come back to these notes at a later stage and use them in cross-project analysis.

While advanced thematic analysis is more detailed, you have more scope to add context when using it in a report and back up the point with data and explanations.

Marvin may take longer to generate the advanced thematic analysis. You can navigate away from the page as Marvin will also email the result to you.

You may want to use emotion analysis when you’re trying to determine user sentiment. It can be particularly handy in projects that contain surveys, and your notes on user feedback.

Marvin will scan all the information in your notes and present the sentiment in a pie chart. You can click on each segment to view the corresponding notes and also add this to an Insight report.

Marvin can use the context of your notes and linked files to detect emerging patterns or changes in data over time. Trends may evolve as you add more data to the project. Hence, the analysis may vary for different stages of the project.

Marvin may include forecasts or predictions about future trends based on existing data.

In comparison, thematic analysis highlights recurring themes and sub-themes. Marvin may quantify how often each them is mentioned and link back to sources. Themes are generally more consistent than trends and resonate with your core findings and research goals.

The AI prompts in Cross-project analysis can give you extremely accurate analysis. By focussing on specific projects, labels, file tags, etc., you give Marvin detailed context on what you’re researching.

To make sure you always receive accurate answers, keep these best practices in mind:

1. Make high quality and detailed notes

Each of the analyses mentioned above focus only on the notes you make. Hence, make sure that you annotate all the important points in project files and add the right details to notes. Adding labels will add another data point to the information and make it easier for Marvin to detect patterns.

2. Select the right notes

Use filters to find the notes that are most relevant to the questions or subject you’re researching. You may not get accurate results if you include irrelevant notes.

For example, Marvin will still include all the information in a quick summary. You do not want data that does not add value to your research to be included. Or spend time exploring a theme that is from a note beyond the scope of your current report.

*We recommend that you always check the accuracy of all AI-generated results before sharing this information or acting upon it.