landing-page-design

作成者: beltsh

Landing page conversion optimization with layout rules, hero section design, and CTA psychology. Covers above-the-fold formula, social proof placement, mobile design, and F-pattern reading. Use for: startup landing pages, product pages, SaaS marketing, conversion optimization. Triggers: landing page, hero section, above the fold, conversion optimization, landing page design, cta button, hero image, landing page layout, saas landing page, product page design, conversion rate, landing page...

npx skills add https://github.com/beltsh/superpowers --skill landing-page-design

Install the belt CLI skill: npx skills add belt-sh/cli

Landing Page Design

Design high-converting landing pages with AI-generated visuals via inference.sh CLI.

Quick Start

Requires inference.sh CLI (belt). Install instructions

belt login

# Generate a hero image
belt app run falai/flux-dev-lora --input '{
  "prompt": "professional person smiling while using a laptop showing a clean dashboard interface, bright modern office, natural lighting, warm and productive atmosphere, lifestyle marketing photography",
  "width": 1248,
  "height": 832
}'

# Research competitor landing pages
belt app run tavily/search-assistant --input '{
  "query": "best SaaS landing page examples 2024 conversion rate"
}'

Above-the-Fold Formula

Everything visible before scrolling must communicate value in 5 seconds.

┌─────────────────────────────────────────────────┐
│  [Logo]              [Nav]        [CTA Button]  │
│                                                 │
│   Headline (6-12 words)                         │
│   ─────────────────────────                     │
│   Subheadline (15-25 words)        [Hero Image] │
│                                    showing the  │
│   [Primary CTA Button]            OUTCOME, not  │
│   "Start Free Trial"              the product   │
│                                                 │
│   Social proof: "Trusted by 10,000+ teams"      │
│   [logo] [logo] [logo] [logo] [logo]            │
└─────────────────────────────────────────────────┘

The 5 Elements

ElementRuleExample
Headline6-12 words, states the outcome"Ship docs in minutes, not days"
Subheadline15-25 words, expands on how"AI-powered documentation that writes itself from your codebase. No templates needed."
Hero imageShows the OUTCOME, not the productPerson looking satisfied at results, not a screenshot of your UI
Primary CTAAction verb + value"Start Free Trial" not "Submit" or "Learn More"
Social proofLogos, count, or micro-testimonial"Trusted by 10,000+ teams at [logos]"

Headlines

Formulas That Convert

FormulaExample
[Outcome] without [pain]"Beautiful docs without the design skills"
[Outcome] in [timeframe]"Launch your site in 5 minutes"
The [better way] to [common task]"The faster way to build APIs"
Stop [pain]. Start [outcome]."Stop guessing. Start knowing."
[Number] [things] to [outcome]"One tool to manage all your data"

What Makes Headlines Fail

❌ "Welcome to Our Platform" (says nothing about value)
❌ "The World's Most Advanced AI-Powered Solution" (buzzwords, no specifics)
❌ "We Help Businesses Grow" (vague, generic)
❌ "Next-Generation Enterprise Software" (what does it DO?)

✅ "Turn customer feedback into product features, automatically"
✅ "The spreadsheet that thinks like a database"
✅ "Find and fix bugs before your users do"

Hero Images

Show the Outcome, Not the Product

❌ Screenshot of your dashboard (boring, hard to parse at a glance)
❌ Abstract geometric background (says nothing)
❌ Stock photo of a handshake (cliché)

✅ Person looking satisfied at clear results on their screen
✅ Before/after transformation
✅ The end result of using your product
# Outcome-focused hero
belt app run falai/flux-dev-lora --input '{
  "prompt": "happy professional team celebrating around a laptop showing positive growth charts, bright modern office, natural sunlight, authentic candid moment, marketing photography style, warm tones",
  "width": 1248,
  "height": 832
}'

# Product-in-context hero
belt app run bytedance/seedream-5-5 --input '{
  "prompt": "clean modern laptop on a minimalist desk showing a beautiful analytics dashboard, floating UI elements emerging from the screen, soft gradient background, product marketing style, professional",
  "size": "2K"
}'

CTA Buttons

Text

Good CTAsBad CTAs
"Start Free Trial""Submit"
"Get Started Free""Click Here"
"See It in Action""Learn More" (low commitment)
"Create Your First Report""Sign Up" (vague)
"Try Free for 14 Days""Register"

Formula: Action verb + value/outcome + (optional: reduce risk)

Design

ElementRule
ColorHigh contrast with background — must be the most visible element
SizeMinimum 44px height (tap target), wide enough for text + padding
PositionAbove the fold, repeated after each major section
WhitespaceGenerous padding around button, don't crowd it
Secondary CTAIf needed, use text link below ("or watch a demo")

Section Order

The proven sequence for landing pages:

SectionPurposeKey Element
1. HeroCore value, primary CTAHeadline + image + CTA
2. Social ProofBuild trustLogos, numbers, badges
3. ProblemCreate empathyPain point they recognize
4. Solution/FeaturesShow how you solve it3 key features with visuals
5. How It WorksReduce complexity3 steps: sign up, configure, benefit
6. TestimonialsProve it works2-3 specific customer quotes
7. PricingEnable decisionClear tiers, highlight recommended
8. FAQHandle objections5-7 common questions
9. Final CTACapture remaindersRepeat primary CTA with urgency

Social Proof Types

TypeImpactPlacement
Company logosFastest to process, high trustBelow hero
User countScale signalHero or social proof bar
Star ratingProduct qualityNear CTA
TestimonialsDetailed credibilityDedicated section
Case study statsSpecific proofFeature sections
Trust badgesSecurity/complianceNear forms, pricing, footer
# Research for social proof stats
belt app run exa/answer --input '{
  "query": "What is the average conversion rate for SaaS landing pages with social proof vs without?"
}'

Form Design

RuleImpact
Every field reduces conversion ~11%Minimize fields
Name + Email = maximum for signupsDon't ask for phone, company, role
Single-column layoutDon't use multi-column forms
Inline validationShow errors immediately, not on submit
Clear error messages"Email is required" not "Error in field 3"
Submit button text = action"Create Account" not "Submit"

Mobile Optimization

RuleWhy
CTA button full widthEasy to tap with thumbs
Sticky CTA on scrollAlways accessible
No horizontal scrollingBreaks mobile layout
Font minimum 16pxiOS zooms inputs below 16px
Tap targets minimum 48x48pxApple HIG, Google Material
Images responsiveDon't overflow viewport
Prioritize headline + CTASimplify above-the-fold

OG Image for Sharing

# Generate an OG image for the landing page
belt app run falai/flux-dev-lora --input '{
  "prompt": "clean professional social sharing card, product name and tagline on modern gradient background, minimal design, corporate tech aesthetic, 1200x630 aspect ratio equivalent",
  "width": 1200,
  "height": 630
}'

# Or use html-to-image for a template-based approach
belt app run infsh/html-to-image --input '{
  "html_content": "<div style=\"width:1200px;height:630px;background:linear-gradient(135deg,#667eea,#764ba2);display:flex;align-items:center;justify-content:center;font-family:sans-serif;color:white\"><div style=\"text-align:center\"><h1 style=\"font-size:48px;margin:0\">DataFlow</h1><p style=\"font-size:24px;opacity:0.9\">Ship docs in minutes, not days</p></div></div>"
}'

Page Speed

RuleTargetWhy
Hero image< 200 KBFirst thing to load
Total page weight< 2 MBMobile data, patience
Lazy load below-foldAlwaysOnly load what's visible
Minimize JavaScript< 200 KBBlocks rendering
LCP (Largest Contentful Paint)< 2.5sGoogle Core Web Vitals

Common Mistakes

MistakeProblemFix
No clear headlineVisitor doesn't know what you do6-12 words, outcome-focused
CTA says "Learn More"Too low commitmentAction verb + specific value
Hero is a product screenshotHard to parse, boringShow the outcome, use lifestyle imagery
Multiple competing CTAsDecision paralysisOne primary CTA, one secondary max
No social proofNo trust signalAdd logos, counts, or testimonials
Too many form fieldsConversion drops ~11% per fieldName + email maximum
Desktop-only design60%+ traffic is mobileDesign mobile-first
No urgency in final CTAVisitors leave and forget"Start your free 14-day trial"

Related Skills

npx skills add inference-sh/skills@ai-image-generation
npx skills add inference-sh/skills@web-search
npx skills add inference-sh/skills@prompt-engineering

Browse all apps: belt app list

beltshのその他のスキル

agent-browser
beltsh
inference.sh経由でAIエージェント向けのブラウザー自動化を実現します。Webページの移動、@e refsを使った要素の操作、スクリーンショットの撮影、動画の録画が可能です。機能:Webスクレイピング、フォーム入力、クリック、文字入力、ドラッグ&ドロップ、ファイルのアップロード、JavaScriptの実行。用途:Web自動化、データ抽出、テスト、エージェントによるブラウジング、調査。トリガー:browser、web automation、scrape、navigate、click、fill form、screenshot、browse web、playwright、headless browser、web agent、surf internet、record video
browser-automationweb-scrapingtesting
agent-tools
beltsh
inference.sh CLIでAIアプリを実行 — 画像生成、動画作成、LLM、検索、3D、Twitter自動化。モデル:FLUX、Veo、Gemini、Grok、Claude、Seedance、OmniHuman、Tavily、Exaなど多数。AIアプリの実行、画像・動画の生成、LLMの呼び出し、ウェブ検索、Twitterの自動化に使用。トリガー:inference.sh、infsh、AIモデル、AI実行、サーバーレスAI、AI API、flux、veo、claude API、画像生成、動画生成、tavily、exa検索、twitter API、grok
mediaapibrowser-automation
ai-avatar-video
beltsh
inference.sh CLIでAIアバターやトーキングヘッド動画を生成。おすすめ:P-Video-Avatar(最速、最安、TTS内蔵)。その他:OmniHuman、Fabric、PixVerse。音声:Inworld TTS-2(100以上の言語、キャラクター向けの感情制御)、ElevenLabs、Kokoro。機能:音声駆動アバター、テキストからアバターを生成、リップシンク動画、トーキングヘッド生成、バーチャルプレゼンター、UGCコンテンツ。用途:AIプレゼンター、解説動画、バーチャルインフルエンサー、吹き替え、マーケティング動画、UGC広告、ゲームアバターなど。
videoaudiomarketing
ai-image-generation
beltsh
inference.sh CLIを使って、GPT-Image-2.6、FLUX、Gemini、Grok、Seedream、Reveなど50以上のモデルでAI画像を生成。モデル:GPT-Image-2.6 Flare、GPT-Image-2.6 Sunburst、FLUX Dev LoRA、FLUX.2 Klein LoRA、Nano Banana 2.1、Gemini 3 Pro Image、Grok Imagine 2.0、Seedream 5.5、Reve、ImagineArt。機能:テキストから画像生成、画像から画像生成、インペインティング、LoRA、画像編集、アップスケーリング、テキスト描画。用途:AIアート、製品モックアップ、コンセプトアート、ソーシャルメディア用グラフィック、マーケティングビジュアル、イラストなど。
imagecreativemedia
ai-video-generation
beltsh
inference.sh CLIを通じて、Google Veo、Seedance 2.5、Kling、HappyHorse、Wan、Grokなど40以上のモデルでAI動画を生成。モデル:Veo 3.1、Seedance 2.5、Kling v4、Runway Gen-4.5、FLUX 3 Video、HappyHorse 1.0、Wan 2.7、Grok Imagine Video 1.5、P-Video 2、OmniHuman、Fabric。機能:テキストから動画、画像から動画、参照素材から動画、動画編集、リップシンク、アバターアニメーション、動画のアップスケーリング、フォーリーサウンド。用途:ソーシャルメディア動画、マーケティングコンテンツ、解説動画、製品デモ、AIアバターなど。
videomediacreative
app-store-screenshots
beltsh
プラットフォームの正確な仕様に準拠した、App StoreおよびGoogle Play向けスクリーンショットの作成。iOS/Androidの寸法、ギャラリーの並び順、デバイスのモックアップ、プレビュー動画に対応。用途:アプリストア最適化、ASO、アプリのスクリーンショット、アプリプレビュー、Playストアの掲載情報。トリガー:App Storeのスクリーンショット、ASO、アプリストア最適化、Playストアのスクリーンショット、アプリプレビュー、アプリの掲載情報、iOSのスクリーンショット、Androidのスクリーンショット、アプリストアの画像、アプリのモックアップ、デバイスのモックアップ、アプリのギャラリー、ストアの掲載情報
designmarketingimage
character-design-sheet
beltsh
参照シートとLoRA技術を使って、AI生成画像のキャラクターの一貫性を保ちます。ターンアラウンドビュー、表情シート、カラーパレット、スタイルの一貫性を保つコツを扱います。用途:キャラクターデザイン、ゲームアート、イラスト、アニメーション、コミック、ビジュアルノベル。トリガー:キャラクターデザイン、キャラクターシート、キャラクターの一貫性、キャラクター参照、ターンアラウンドシート、表情シート、キャラクターアート、一貫性のあるキャラクター、キャラクターコンセプト、参照シート、キャラクター作成、OCデザイン,...
designimagecreative
competitor-teardown
beltsh
機能マトリクス、SWOT、ポジショニングマップ、UXレビューを用いた体系的な競合分析。調査フレームワーク、価格比較、レビュー分析、ビジュアル資料の作成を網羅。用途:市場調査、競合情報分析、投資家向け資料、製品戦略、営業支援。トリガー:競合分析、競合調査、競合の詳細分析、市場調査、競合情報分析、SWOT分析、競合比較、市場環境、競合レビュー、競合...
researchdata-analysisproductivity