Documentation

The Generation Process

1. Input Collection: You provide client information (name, website, Google Ads data, target audience)
2. Context Building: The system scrapes the client's website to extract brand colors, typography, and key messaging
3. Prompt Customization: Your target persona, industry, and key objections are injected into the active AI prompt template
4. AI Generation: The AI generates a complete landing page with hero section, features, testimonials, and CTA forms
5. Preview & Edit: You can preview the generated page and make real-time edits to text, colors, and layout
6. Export: Download the landing page as a standalone package ready for Vercel deployment

Key Features

• Message Match Optimization: Headlines automatically match your Google Ads copy for better Quality Score
• Brand Consistency: Colors and typography extracted from the client's existing website
• B2B Best Practices: Copy follows proven conversion frameworks (metric-driven, objection-handling, specificity)
• Real-time Editing: Visual editor allows instant changes without regenerating the entire page

Client Information

• Client Name (Required): The company name that will appear throughout the landing page
• Website URL (Optional): If provided, we'll analyze their site to extract brand colors, fonts, and messaging. This ensures the generated page matches their existing brand identity.

Onboarding Form Upload

Upload a PDF or Word document containing client onboarding information. The AI will extract:

• Target audience and buyer personas
• Pain points and challenges
• Value propositions and key benefits
• Competitive differentiators

Google Ads Campaign Data

• Target Keywords: Enter the keywords you're bidding on (comma-separated). These will be incorporated into the page copy for SEO and message match.
• Ad Copy/Headline: Paste your actual Google Ads headline. The landing page headline will mirror this to improve Quality Score and reduce bounce rate.

Target Audience & Messaging

• Target Persona: Select the primary decision-maker (CFO, CTO, VP Sales, etc.). This customizes the AI prompt to use language and pain points specific to that role.
• Industry: Choose the client's industry (SaaS, Consulting, E-commerce, etc.). The AI will use industry-specific examples and terminology.
• Key Objections: List the main concerns this persona typically has (e.g., "Too expensive, implementation takes too long"). The AI will proactively address these objections in the copy.

What are Prompt Settings?

The Prompt Settings page allows you to view and edit the AI instructions that generate landing pages. Think of it as the "recipe" that tells the AI how to write compelling B2B copy.

You can access Prompt Settings from the main dashboard by clicking the "Prompt Settings" button in the header.

How the Prompt Generator Works

The prompt generator is the "brain" behind the landing page creation. Here's the exact workflow:

1. You fill out the generator form with client name, website URL, Google Ads data, target persona, industry, and key objections.
2. The system loads the active prompt template from the database (you can view/edit this in Prompt Settings).
3. Dynamic variables are injected: The template contains placeholders like {target_persona}, {industry}, and {key_objections}. These are replaced with your actual form values.
   Example:
   Template: "Write copy for {target_persona} in the {industry} industry."
   After injection: "Write copy for CFO in the SaaS industry."
4. Website scraping (if URL provided): The system analyzes the client's website to extract brand colors, fonts, and key messaging. This data is added to the prompt context.
5. Onboarding form parsing (if uploaded): The AI reads the PDF/Word document and extracts pain points, value props, and differentiators.
6. The complete prompt is sent to the AI: The AI receives the system prompt (with adversarial checks and quality gates) + your customized variables + scraped brand data + onboarding insights.
7. The AI generates the landing page: It creates hero section, features, testimonials, FAQ, and CTA forms following the prompt instructions.
8. The output is saved to the database with a unique slug, and you're redirected to the preview page where you can edit and export.

Key Components

• System Prompt: The core instructions that define the AI's writing style, tone, and structure. This includes:
  • Adversarial checks (Ruthless Competitor, Cynical Consumer, Distracted Scroller)
  • Quality gates (Objection Gauntlet, Specificity Audit, Voice Authenticity Test)
  • Copywriting rules (headline formulas, CTA patterns, testimonial structure)
  • Google Ads best practices (message match, Quality Score optimization)
• Dynamic Variables: Placeholders that get replaced with actual values from the generator form:
  • {target_persona} → CFO, CTO, VP Sales, Marketing Director, etc.
  • {industry} → SaaS, Tech, AI, Professional Services
  • {key_objections} → "Too expensive, implementation takes too long, ROI unclear"
  • {client_name} → The company name
  • {google_ads_headline} → The actual ad copy for message match
• Model Configuration: Temperature, max tokens, and other AI parameters that control creativity vs. consistency. You can adjust these in Prompt Settings.

What to Put in Each Field

Editing Prompts

1. Navigate to Prompt Settings from the dashboard
2. Click Edit Prompt to modify the active prompt
3. Make your changes in the editor (use {variable_name} syntax for dynamic fields)
4. Click Save as New Version to create a new prompt version
5. The new version automatically becomes active and will be used for all future generations

Version History

Every time you save a prompt, a new version is created. This allows you to:

• Track changes over time
• Roll back to previous versions if needed
• Compare performance between different prompt versions
• Maintain a library of specialized prompts for different use cases

Testing Prompts

Use the Prompt Comparison feature to test different prompts side-by-side:

1. Click Test Prompt from the Prompt Settings page
2. Enter test client information
3. Generate pages with different prompt versions
4. Compare the output quality, tone, and conversion elements
5. Activate the best-performing prompt

Input Quality

• Always provide a website URL when available—brand color extraction dramatically improves visual consistency
• Use actual Google Ads copy instead of generic headlines—message match is critical for Quality Score
• Be specific with objections—"Implementation takes 6+ months" is better than "Too complex"
• Upload detailed onboarding forms—the more context the AI has, the more personalized the output

Editing Generated Pages

• Start with text edits before changing colors—the AI usually gets the copy structure right
• Keep headlines under 10 words—shorter headlines convert better on mobile
• Maintain metric specificity—if you edit copy, keep the numbers and timeframes
• Test CTAs—try "Get Your Free Demo" vs. "See How It Works" and track which performs better

Prompt Optimization

• Track which prompts generate the best pages—use the version history to identify patterns
• Create industry-specific prompts—SaaS copy differs from consulting or e-commerce
• A/B test prompt changes—generate two versions and see which converts better
• Update prompts quarterly—B2B copywriting trends evolve, and your prompts should too

Common Pitfalls to Avoid

• Don't skip the target persona field—generic copy converts 40% worse than persona-specific copy
• Don't use vague objections—"Price" is too broad; "Costs 3x more than current solution" is actionable
• Don't ignore brand colors—if the website scraper finds colors, use them (or the page will look off-brand)
• Don't over-edit—the AI follows proven B2B frameworks; trust the structure and tweak the details