Custom studies are chat-based research formats you can design for specific use cases, like message testing, focus groups, reputation reactions, or competitive analysis. Once created, they can be launched anytime using a simple command in chat.
Tip: Custom studies are ideal for reusable workflows. You can create one for each type of testing your team does regularly, then run them anytime, with any of your AI-lookalike Audiences.
Step 1: Go to Custom Studies
Click Custom Studies in the sidebar
Click + Create Custom Study in the top-right corner
This opens a two-step setup process.
Step 2: Name and Describe the Study
What should we call this custom study?
Give it a clear name your team will recognize. This name will appear in the Custom Studies list.
What does this custom study do?
Briefly describe the purpose of the study—what it helps you test or evaluate, and what kind of outputs it provides.
The command name
This is the shortcut your team will type in chat (e.g. /reputation-reactions or /hook-testing). It auto-generates from the study name, but you can edit it.
Click Next: Set up questions to continue.
Step 3: Set Up Study Inputs
Now define the questions Gutsy should ask when someone runs the study in chat. These ensure Gutsy collects the right information to make accurate predictions.
For each question:
Use plain, conversational phrasing (e.g. “What persona(s) do you want to use?”)
Toggle Required question on or off depending on whether it’s essential
Click Add another question to include additional fields
You can also click Advanced options to add notes or internal guidance, if needed.
When you're done, click Save changes.
Step 4: Launch in Chat
Your custom study will now appear in the Custom Studies tab, with its command visible at the top of the card.
To run it:
Go to the Chat tab
Type / and select the custom study by name
Fill in the answers as prompted
Gutsy will generate predictions based on your inputs
Best Practices
Choose Clear Names
Good: "Customer Feedback Analysis", "Competitor Pricing Review"
Avoid: "Tool 1", "AI Helper", "Analysis Thing"
Write Descriptive Explanations
Good: "Analyzes customer reviews to identify common pain points, feature requests, and satisfaction trends. Provides actionable insights for product and service improvements."
Avoid: "Looks at feedback", "Helps with customers"
Describe the Output your Expect
Add formatting preferences: "Present findings as numbered list"
Set tone: "Write in executive summary style"
Include requirements: "Always end with 3 specific recommendations"
Craft Specific Questions
Good: "What is the URL or name of the competitor you want to analyze?"
Avoid: "Competitor?", "What do you want?"