Skip to main content
Dialog’s research chat is the core of the product. Here’s how to get the most out of it.

​
Writing a good query

The best queries are specific about what you want to learn. Here are some patterns that work well:

Specific and focused

“What do SaaS founders on Reddit say about pricing their products? What strategies work and what doesn’t?”

Comparative

“How do people compare Notion vs Obsidian for personal knowledge management? What are the main pros and cons of each?”

Sentiment-focused

“What’s the general sentiment around AI coding assistants? Are developers excited or skeptical?”

Discovery-oriented

“What subreddits discuss indie game development? What are the most active communities?”

​
Understanding results

When Dialog processes your query, you’ll see several elements in the response:

​
Thinking indicators

Dialog shows its reasoning process in real time. You can see which communities it’s searching, what tools it’s using, and how it’s analyzing the data.

​
Source citations

Every insight includes links back to the original Reddit posts and comments. Click any citation to view the source material directly.

​
Structured findings

Results are organized into clear sections with headers, bullet points, and sometimes tables. Dialog identifies themes, patterns, and notable outliers in the data.

​
Follow-up questions

You can ask follow-up questions in the same session to dig deeper. Dialog remembers the context of your conversation: 
Initial: "What are people saying about Vercel on Reddit?"
Follow-up: "What about their pricing specifically? Any complaints?"
Follow-up: "How does that compare to Netlify pricing sentiment?"

​
Session management

  • Sessions are automatically saved as you go
  • Switch between sessions using the sidebar
  • Each session has a generated title based on your first question
  • You can delete sessions you no longer need