Project Setup

Step 1: Analyze your business automatically

Quick Start

The fastest way to get started is with automatic domain analysis:

1. Enter your website domain (e.g., loebermercedes.com)
2. Click "Auto-Fill Project Details"
3. Our AI will analyze your website and extract:
   • Brand name and industry
   • Key competitors
   • Relevant search topics
   • Target audience
4. Select which topics you want to test
5. Click "Create Project"

Example Analysis

For a Mercedes dealership website, our AI automatically detects:

Manual Setup

You can also create projects manually by filling out each field individually if you prefer more control over the analysis.

Prompt Generation

Step 2: Create natural search queries

What Are Prompts?

Prompts are the questions and queries that customers naturally ask when researching your type of business. ChatViz generates realistic queries that people might ask AI assistants.

Types of Queries Generated

Comparison Queries

• "Mercedes GLS vs BMW X7 — which is better?"
• "Compare Asana vs Monday.com for project management"
• "Tesla Model Y vs Mercedes EQC electric SUV"

Best-of Queries

• "Best luxury SUV for families in Chicago"
• "Top project management software for remote teams"
• "Best CRM for small businesses"

Decision-Making Queries

• "Should I buy or lease a Mercedes GLS?"
• "Is Asana worth it for small teams?"
• "What's the most reliable luxury car brand?"

Feature-Specific Queries

• "Mercedes with best safety features"
• "Project management with time tracking"
• "CRM with email marketing integration"

Customizing Queries

You can influence the types of queries generated by adding specific rules:

Best Practices

• Be Customer-Focused — Think about real questions your customers ask
• Include Location — Local searches are very important
• Cover the Funnel — From awareness to decision-making queries
• Include Competitors — People often compare options

Running Prompts

Step 3: Test against AI models

AI Model Selection

Choose which AI models to test your queries against:

• Claude Sonnet 4.6 — Balanced performance and cost
• GPT-4o Mini — Fast and cost-effective
• Gemini 2.5 Flash — Google's efficient model

Running the Analysis

1. Select your prompt set from the dropdown
2. Choose which AI model to test
3. Click "Run Against AI"
4. Watch the progress in the debug output
5. Review results when complete

What to Look For

Positive Signals

• Brand Mentions — Your business is specifically named
• Recommendations — AI suggests your product/service
• Accurate Information — AI provides correct details about your business
• Prominent Placement — You're mentioned early in the response

Areas for Improvement

• No Mention — Your business doesn't appear in responses
• Competitor Emphasis — Competitors are mentioned instead
• Incorrect Information — AI has wrong details about your business
• Generic Responses — AI gives general advice without specific recommendations

Interpreting Results

Use the results to understand:

• How visible your brand is to AI models
• What information AI models associate with your business
• How you compare to competitors in AI responses
• Which types of queries work best for your business

Examples

Real-world use cases and results

Mercedes Dealership

Domain: loebermercedes.com

Generated Topics

• "best luxury SUV Chicago"
• "Mercedes GLS vs BMW X7"
• "Mercedes dealership near me"
• "certified pre-owned luxury cars"

Sample Results

SaaS Company

Domain: asana.com

Generated Topics

• "best project management for remote teams"
• "Asana vs Monday.com vs Notion"
• "project management with time tracking"
• "team collaboration software"

Sample Results

Best Practices

Tips for better AI visibility

Content Strategy

• Create Comprehensive Pages — AI models prefer detailed, authoritative content
• Use Natural Language — Write how customers actually speak
• Include Comparisons — Help AI understand your competitive position
• Add Local Context — Include location-specific information
• Update Regularly — Fresh content signals authority to AI models

SEO for AI

• Structured Data — Use schema markup to help AI understand your content
• Clear Headings — Organize content with descriptive headings
• Answer Questions — Create FAQ sections addressing common queries
• Feature Benefits — Clearly explain what makes you different

Testing Strategy

• Test Regularly — AI models update frequently
• Monitor Competitors — See how they perform in AI responses
• Vary Queries — Test different ways customers might ask
• Track Changes — Monitor how your visibility changes over time

Troubleshooting

Common issues and solutions

Domain Analysis Issues

Auto-fill not working

• Make sure you entered just the domain (e.g., "example.com" not "https://example.com")
• Check that your website is accessible and has sufficient content
• Try the manual setup if automatic analysis fails

Incorrect business details detected

• You can edit any auto-filled information before creating the project
• Make sure your website clearly describes your business and industry
• Consider adding more descriptive content to your homepage

Prompt Generation Issues

Irrelevant prompts generated

• Refine your rules to be more specific about your target audience
• Include industry-specific terminology
• Add location context if you serve local customers

Not enough prompts created

• Try broader topics or add more variety to your rules
• Include different types of queries (comparison, best-of, decision-making)
• Consider your customer journey from awareness to purchase

AI Response Issues

No brand mentions

• Your website might not have enough authority or visibility
• Try more specific, local queries
• Improve your website's content depth and quality
• Consider if your brand is well-known enough for general queries

Incorrect information in AI responses

• Ensure your website has accurate, up-to-date information
• Submit corrections to major directories and review sites
• AI models may have outdated training data

Getting Help

If you're still having issues:

• Check the debug output for error messages
• Try simplifying your queries to test basic functionality
• Ensure your website is publicly accessible
• Contact support with specific error messages or unexpected behavior