Overview
Your OneGlance workspace tracks how AI providers mention your brand. This guide walks you through creating a workspace, configuring your brand details, and running your first analysis.Each workspace tracks one brand across multiple AI providers (ChatGPT, Claude, Perplexity, Gemini, etc.). If you track multiple brands, create separate workspaces.
Prerequisites
Before starting, you’ll need:- A OneGlance account
- Your brand name (e.g., “Pipedrive”)
- Your brand’s primary domain (e.g., “pipedrive.com”)
Step 1: Create Your Workspace
Navigate to Workspace Creation
After logging in, click Create Workspace from the sidebar or dashboard.
Enter Brand Details
Fill in the workspace creation form:
- Brand Name: The name AI providers should mention (e.g., “HubSpot”, “Salesforce”)
- Brand Domain: Your primary website domain without protocol (e.g., “hubspot.com”)
- Organization Name (optional): Your company name if different from brand
Step 2: Configure Your First Prompts
Prompts are the questions OneGlance asks AI providers on your behalf. Good prompts reflect real buyer questions in your category.Understanding Prompt Strategy
Effective prompts:- Mirror buyer intent: Use questions your prospects actually ask
- Cover key use cases: Include prompts for different buyer personas and scenarios
- Test competitive positioning: Ask questions where competitors might surface
- Vary specificity: Mix broad category queries with specific feature comparisons
Adding Prompts During Onboarding
Use Suggested Prompts
OneGlance provides category-appropriate suggestions. Click any suggestion to add it to your prompt list.Example suggestions for a CRM brand:
- “What are the best alternatives to for growing teams?”
- “Which tools are most recommended for sales pipeline management?”
- “Compare with top competitors for pricing and value.”
What Happens During First Run
The onboarding run executes in this sequence:-
Job Submission (apps/web → packages/services)
-
Provider Queue Distribution (packages/services/src/agent/jobs.ts:59)
- Parallel Execution: Each provider (ChatGPT, Claude, etc.) processes all prompts simultaneously
- Response Storage: Raw responses are saved to ClickHouse
-
AI Analysis: OpenAI analyzes each response for:
- Brand mentions and sentiment
- Competitive positioning
- GEO score (Generative Engine Optimization)
- Citations and sources
Monitoring Progress
Watch the real-time progress tracker:Provider Status Indicators
Provider Status Indicators
- Pending: Job queued but not started
- Running: Provider actively processing prompts (spinning loader)
- Completed: All prompts processed successfully (green checkmark)
- Failed: Provider encountered an error
apps/web/src/app/(auth)/onboarding/page.tsx:293:Step 3: Configure AI Provider Settings
By default, OneGlance queries all supported AI providers. You can customize this in Settings.Edit Active Providers
In the AI Providers section, click the edit icon (pencil).Select which providers to query:
- ChatGPT (OpenAI GPT-4)
- Claude (Anthropic)
- Perplexity (Perplexity AI)
- Gemini (Google)
- MetaAI (Meta)
When to Adjust Providers
Reduce costs: Fewer providers = fewer API calls Focus on key channels: If your buyers primarily use ChatGPT, prioritize it Expand coverage: Enable all providers for comprehensive visibility trackingStep 4: Understanding Your Dashboard
After your first run completes, the dashboard displays:Key Metrics
Presence Rate
Percentage of responses where your brand was mentioned.Example: 65% presence means your brand appeared in 13 of 20 AI responses.
Average Rank
Your brand’s average position when listed alongside competitors.Example: Rank #2 means you’re typically the second brand mentioned.
Top Source
Most frequently cited domain in AI responses about your brand.Example: “g2.com” indicates G2 reviews heavily influence AI recommendations.
Top Competitor
Brand most often mentioned alongside yours.Example: “HubSpot” appearing in 80% of responses where you’re mentioned.
GEO Score Explained
The GEO (Generative Engine Optimization) Score (0-100) measures your brand’s AI visibility using four weighted components:- Visibility (25%): How prominently your brand appears (placement, coverage, frequency)
- Rank (25%): Position in recommendation lists (#1 = 100, #2 = 80, etc.)
- Sentiment (25%): Tone of mentions (positive/neutral/negative)
- Recommendation (25%): Endorsement strength (top_pick = 100, conditional = 60, etc.)
GEO Score Interpretation
GEO Score Interpretation
- 80-100: Dominant - Brand is a top recommendation with strong positioning
- 60-79: Strong - Consistently mentioned with favorable sentiment
- 40-59: Moderate - Present but not strongly recommended
- 20-39: Weak - Mentioned in passing or with caveats
- 0-19: Absent/Minimal - Rarely appears or negatively positioned
Step 5: Invite Team Members
Add Members by Email
In the Members section:
- Enter teammate’s email
- Select role (Member or Owner)
- Click Add
Role Permissions
| Permission | Member | Owner |
|---|---|---|
| View dashboard & prompts | ✅ | ✅ |
| Run prompts & analysis | ✅ | ✅ |
| Edit prompts & schedule | ✅ | ✅ |
| Add/remove members | ❌ | ✅ |
| Edit brand details | ❌ | ✅ |
| Delete workspace | ❌ | ✅ |
Troubleshooting
First run shows no data after 10+ minutes
First run shows no data after 10+ minutes
Possible causes:
- All providers failed (check status indicators on onboarding page)
- No prompts were saved (verify on Prompts page after redirecting)
- Network issues during job submission
- Go to Schedule page
- Click “Save Schedule” (this triggers an immediate run)
- Monitor progress on Dashboard
Brand domain not detecting mentions correctly
Brand domain not detecting mentions correctly
Common mistakes:
- Including
https://orwww. - Using a subdomain (use
company.com, notapp.company.com) - Using a redirect domain
GEO scores seem low for my brand
GEO scores seem low for my brand
This is often accurate, not a bug. AI providers may:
- Not recognize your brand in your category
- Prioritize competitors with more public content
- Lack recent information about your product
- Review Interpreting Metrics to understand scoring
- Check Managing Prompts to refine your prompt strategy
- Analyze which sources are cited—these influence AI recommendations
Workspace code not working for teammates
Workspace code not working for teammates
Possible causes:
- Teammate using an organization invite code instead of workspace code
- Code copied incorrectly (check for extra spaces)
Next Steps
Managing Prompts
Learn to refine prompts, add variations, and optimize for better visibility
Scheduling
Automate regular runs to track AI visibility trends over time
Interpreting Metrics
Deep dive into GEO scores, sentiment analysis, and competitive positioning
Team Collaboration
Best practices for multi-user workflows and access control
Related Concepts
- Quickstart - Fast setup overview
- API Reference - Programmatic access to workspace data
- Local Setup - Self-hosting and custom deployments