The Best Claude Prompts for Landing Page Optimization

Claude stands out from other language models for a precise reason when it comes to landing page optimization: its ability to reason in a structured way about persuasion problems, information architecture, and consumer psychology simultaneously. Unlike more generic tools, Claude maintains logical coherence across copy sections, identifies contradictions in value messaging, and can alternate perspectives — those of the copywriter, CRO specialist, and end user — within a single response. This makes it an exceptionally powerful tool for anyone working on landing page optimization.

In this guide you'll find prompts tested and refined directly on Claude (versions 3 Sonnet and Opus), designed to cover every phase of the optimization process: from critical analysis of existing copy, to generating A/B headline variations, to structuring visual hierarchy and optimizing CTAs. Each prompt comes with an explanation of the underlying prompt engineering principles, guidance on expected output, and personalization tips.

Whether you're optimizing a landing page for B2B SaaS, e-commerce, or lead generation campaigns, the frameworks presented here are immediately applicable. The goal isn't just to provide you with prompts to copy, but to transfer the compositional logic that allows you to build your own prompts independently and systematically.

Ready-to-Use Prompts for Landing Page Optimization

1. Critical Analysis of Existing Copy

You are a CRO (Conversion Rate Optimization) expert with 10 years of experience analyzing B2B and B2C landing pages. Analyze the following landing page text using these criteria:

1. Clarity of value proposition: is it immediately understandable within 5 seconds?
2. Alignment between headline and sub-headline
3. Social proof and credibility (testimonials, data, reviews)
4. Strength and specificity of CTAs
5. Unaddressed objections that could block conversion
6. Message coherence relative to search intent [insert primary keyword]

For each criterion provide: rating (1-5), specific problem identified, concrete solution with reformulated example.

LANDING PAGE TEXT:
[paste your text here]

TARGET AUDIENCE:
[describe your ICP: role, industry, primary pain point]

Why this prompt works: It assigns Claude a specific role with years of experience (role prompting), defines numeric evaluation criteria that force structured rather than vague responses, and includes the "5-second" constraint — a real CRO benchmark that orients the model toward concrete professional standards. The 1-5 rating request per criterion prevents generic answers.

Expected output: A structured table or list with numeric ratings, precise diagnoses, and alternative copy reformulations. Claude tends to produce 600-900 words of detailed analysis with this framework.

Personalization tip: Add "PRIMARY COMPETITOR: [URL or description]" at the end to also request competitive positioning analysis against your main competitors.

2. Headline Generation for A/B Testing

Generate 10 headline variations for a landing page with these specifications:

PRODUCT/SERVICE: [2-line description]
AUDIENCE: [detailed ICP]
PAGE OBJECTIVE: [lead gen / direct purchase / demo request]
PRIMARY SEO KEYWORD: [insert keyword]
BRAND TONE: [e.g., authoritative, friendly, technical, provocative]

For each headline:
- Write the main headline (max 8 words)
- Write a supporting sub-headline (max 20 words)
- Indicate the psychological lever used (e.g., FOMO, numeric specificity, direct benefit, social proof, curiosity gap)
- Indicate which audience segment it works best for

Distribute the 10 variations across these frameworks:
- 3 headlines focused on final benefit
- 3 headlines focused on problem resolution
- 2 provocative / counter-intuitive headlines
- 2 headlines with numeric specificity

Why this prompt works: The distribution constraint (3+3+2+2) prevents Claude from producing 10 homogeneous variations — a common mistake when simply asking "generate alternative headlines." The explicit request for psychological levers transforms the output into educational material beyond operational use, allowing you to select variations with strategic awareness.

Expected output: A formatted list with 10 headline/sub-headline pairs, each accompanied by a psychological lever label and a note on target segment. Ideal for planning A/B tests with clear hypotheses.

Personalization tip: Replace the 3+3+2+2 distribution with frameworks specific to your industry. For enterprise SaaS, you might request 4 ROI-oriented variations and 3 that address security/compliance objections.

3. Hero Section Optimization

Analyze and rewrite the hero section of this landing page following the AIDA structure adapted for modern web:

TARGET STRUCTURE:
- Headline (problem or transformation in 6-8 words)
- Sub-headline (benefit amplification with specificity, 15-25 words)
- Bullet points of value (3 points, format "action verb + specific benefit + quantification")
- Primary CTA (button text + supporting microcopy below button)
- Risk reduction element (guarantee, free trial, no credit card required, etc.)

CURRENT HERO TEXT:
[paste here]

CONSTRAINTS:
- Keep the SEO keyword "[insert keyword]" in the headline or sub-headline
- Tone should be [formal/informal/technical]
- Avoid these commonly overused words/phrases in the industry: [e.g., "innovative solution," "industry leader," "best-in-class"]

After the rewrite, explain in 3 points the copywriting choices you made.

Why this prompt works: It provides a detailed output structure that Claude must follow (structural prompting), includes explicit negative constraints (words to avoid) that are often more effective than positive constraints, and requests a meta-explanation that forces Claude to make explicit its choices — useful for learning and for defending decisions with internal stakeholders.

Expected output: A complete, implementation-ready hero section followed by a brief analysis of strategic choices. The meta-explanation request produces higher-quality output because Claude "thinks aloud" before writing.

Personalization tip: Add "HEATMAP DATA: [describe where users click or abandon]" if you have access to Hotjar or Microsoft Clarity data — Claude integrates this behavioral data into its analysis.

4. Page Structure and Content Hierarchy

You are an information architecture and UX writing expert for high-conversion landing pages. Based on the following inputs, design the complete structure of a landing page:

OBJECTIVE: [e.g., generate demo requests for HR software]
PRIMARY TRAFFIC: [e.g., Google Ads with keyword "attendance management software"]
AWARENESS STAGE: [e.g., medium — user knows the problem but not the solution]
PRINCIPAL OBJECTIONS: [list 3-5 real objections from interviews or reviews]
MAIN COMPETITORS: [names or descriptions]

Provide:
1. Ordered list of page sections with title, objective of each section, and recommended content type
2. For each section: indicate the implicit user question it answers
3. Recommended CTA positioning (primary and secondary) with reasoning
4. Trust elements to include (type and position)
5. Optimal structure for mobile-first (indicate which sections to shorten or reorder on mobile)

Why this prompt works: The "awareness stage" parameter is drawn from Eugene Schwartz's framework on market consciousness levels — a technical concept that orients Claude toward strategic decisions on the depth of education needed. Including real objections (ideally extracted from G2, Capterra reviews, or interviews) transforms the prompt from a theoretical exercise into data-driven design tool.

Expected output: A detailed landing page roadmap with 7-12 sections, each with declared purpose and implicit user question. Particularly useful for briefing designers and developers.

Personalization tip: Directly paste 5-10 real product reviews (or competitor reviews) in the "PRINCIPAL OBJECTIONS" section — Claude autonomously extracts recurring patterns and transforms them into structured objections.

5. Call-to-Action Optimization

Analyze and optimize all CTA elements on this landing page. For each existing CTA:

1. Evaluate the effectiveness of button text (specific vs generic, value-oriented vs action-oriented)
2. Analyze supporting microcopy (present/absent, reduces or increases friction?)
3. Verify coherence between CTA promise and the following page/step
4. Propose 3 alternative variations tested against cognitive friction reduction principles

Subsequently, generate a CTA matrix for different points on the page:
- CTA above the fold (cold user, first impression)
- CTA after benefits section (interested user)
- CTA after testimonials (evaluating user)
- CTA exit-intent or footer (last chance)

For each CTA in the matrix: button text + microcopy + urgency/scarcity element if appropriate.

CURRENT CTA TEXT:
[paste all present CTAs here]

POST-CLICK ACTION:
[e.g., 3-field form, immediate call, PDF download, payment]

Why this prompt works: The matrix by page position is an advanced CRO technique that considers user warmth level (cold vs warm traffic) — a parameter most generic prompts ignore. The explicit request for microcopy recognizes that text below the button is often more determinative than button text itself in reducing friction.

Expected output: Analysis of existing CTAs with scoring and alternatives, plus a complete CTA matrix optimized by position. Tabular or structured list format with approximately 400-600 words.

Personalization tip: Add "CURRENT CONVERSION RATE: [X%]" and "INDUSTRY BENCHMARK: [Y%]" to contextualize the analysis — Claude calibrates the urgency of recommended changes based on the gap to be closed.

6. Testimonial Analysis and Rewriting

The following testimonials are currently on my landing page. For each:

1. Rate persuasive effectiveness (specificity, credibility, relevance to target)
2. Identify missing elements (numeric results, before-problem context, customer role/company)
3. Propose a rewritten version that maintains authenticity but maximizes persuasive impact
4. Indicate which specific target objection this testimonial should address

After individual analysis:
- Suggest the optimal order for presenting testimonials
- Indicate which testimonial types are missing to cover the entire buyer decision cycle
- Provide 5 questions to ask customers for gathering high-impact testimonials

CURRENT TESTIMONIALS:
[paste testimonials with name, role, company if available]

PRINCIPAL OBJECTIONS:
[list 3-5 objections]

Why this prompt works: It explicitly links each testimonial to a specific objection — a strategic social proof technique often overlooked. The request for questions to gather new testimonials transforms the prompt from an optimization tool into a content production tool, multiplying its value over time.

Expected output: Structured analysis for each testimonial, rewritten versions, and a list of 5 interview questions. Output typically 700-1,000 words.

Personalization tip: Include a detailed ICP profile — Claude will select which testimonial aspects resonate most with that specific profile and calibrate rewrites accordingly.

Prompt Engineering Techniques for Landing Page Optimization

1. Use the "Awareness Stage" Parameter to Calibrate Depth

The concept of market awareness (from Eugene Schwartz) is one of the most powerful parameters you can provide Claude. Specify whether your target user is:

• Unaware: doesn't know they have the problem
• Problem Aware: knows the problem but not the solutions
• Solution Aware: knows the solutions but not your product
• Product Aware: knows your product but isn't convinced
• Most Aware: is already convinced, just needs an offer

Claude will significantly adapt tone, educational depth, and argument structure based on this parameter. A "Solution Aware" user requires copy that differentiates your product, not copy that explains why the problem exists.