landing-page-funnel
Skill3 starsChanged 46 days agoMITView on GitHub
Use when auditing, writing or testing landing pages for conversion — above-the-fold structure, ad-to-page message match, body section order, form optimization, social proof, objection handling, mobile and page speed (Core Web Vitals), and A/B test design with proper sample sizes. Also use when someone says "traffic is fine but nobody converts", asks for a landing page teardown, or needs CRO test ideas. Don't use for full website copywriting, e-commerce product detail pages, blog/SEO content (geo-ai-seo) or tracking bugs (gtm-debugging).
1 other project uses a version of this.
Adopted by
Most recently taken up by
- puretechnyc/purebrain-skillstheir own version, first seen May 24, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- puretechnyc/purebrain-skills★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository14 other files it gives its agents.
- brandingSkill
- competitor-researchSkill
- geo-ai-seoSkill
- google-ads-diagnosticsSkill
- google-ads-keywordsSkill
- gtm-debuggingSkill
- instagram-organicSkill
- linkedin-organicSkill
- market-positioningSkill
- meta-ads-creativeSkill
- meta-ads-diagnosticsSkill
- pinterest-organicSkill
- tiktok-adsSkill
- youtube-organicSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 30
Don't use for: multi-page website copy, e-commerce product detail pages and checkout (different model), blog or AI-search content (use …
- Distinctive: line 34
Ask for (or extract from the conversation):
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Landing Page Funnel Optimization
Overview
Audit and improve landing pages so paid and organic traffic converts: match the page to the ad that brought the visitor, answer "what, why me, what next" within seconds, order the body to build belief, remove friction from the form, and prove claims with specific social proof. Includes a test design framework so improvements are measured rather than assumed.
When to Use
- Auditing an existing landing page that gets traffic but few conversions
- Writing or rewriting landing page copy for a campaign
- Designing A/B tests and prioritizing what to test first
- Aligning landing pages with ad creative (message match)
- Optimizing lead forms, trial sign-ups or booking flows
- Choosing the right page type for each funnel stage
Don't use for: multi-page website copy, e-commerce product detail pages and checkout (different model), blog or AI-search content (use geo-ai-seo), or diagnosing why conversions are not recorded (use gtm-debugging).
Inputs
Ask for (or extract from the conversation):
- The page: URL (fetch it if you can) or pasted copy plus screenshots of desktop and mobile
- Traffic source and the ad/email that sends visitors (headline, offer, visual)
- Conversion goal and current conversion rate, visitors per week, device split
- Audience and offer: who, what they get, price or commitment, guarantee
- Proof available: testimonials, numbers, logos, reviews, certifications
- Form fields currently required, and which are actually needed to fulfil
- Speed data: PageSpeed Insights / Core Web Vitals if available
- Analytics: scroll depth, form abandonment, heatmaps if available
Data rules: if you cannot fetch the page, say so and work from what the user provides; do not describe elements you have not seen. Never invent testimonials or numbers — use [insert proof]. Benchmarks are starting points; the page's own history and the business's unit economics decide what "good" is. Use the account's currency.
Landing Page Anatomy
Above the fold (first screen on mobile and desktop)
Within ~5 seconds the visitor must know what this is, why it matters to them, and what to do next.
| Element | Rule |
| H1 | Restates the ad's promise in the visitor's words. Word-for-word match with the ad headline is a feature. |
| Sub-headline | One sentence with a specific benefit, number or mechanism |
| Hero visual | The product or outcome in use; no abstract stock imagery |
| Primary CTA | One action, one button, visible without scrolling on mobile; verb-first label |
| Risk reducer near CTA | "No credit card", "Cancel anytime", "Free 15-min call", a star rating with count |
| Navigation | Removed or minimal; every exit link is a leak |
Body order
- Problem — name the pain in the visitor's language (2–3 sentences)
- Proof of understanding — a data point, example or quote that shows you get it
- Solution — the transformation, not the feature list
- How it works — 3 steps, visual
- Social proof — specific testimonials, logos, results
- Objections — the 2–3 real ones (price, time, risk, "will it work for me")
- Offer recap and CTA — the same action, plus a lower-commitment alternative
- FAQ — remaining objections in question form (also helps AI and search extraction)
Form optimization
| Factor | Rule |
| Fields | Ask only what is needed to deliver the next step; every extra required field costs conversions. Enrich later. |
| Multi-step | For 5+ fields, split into steps starting with the easiest (interest), ending with contact details; show "Step 1 of 3" |
| Validation | Inline, as the user types; message on the field, not a banner |
| Defaults and autofill | Correct autocomplete attributes; prefill country/timezone |
| Button label | The outcome: "Get my quote", "Start free trial", "Book the call" — never "Submit" |
| Privacy line | One sentence under the button: what happens with the data, no spam |
| Fallback | A lower-commitment option under the form: "Not ready? Get the guide" |
| Mobile | Tap targets ≥ 44×44 px; correct keyboard types (type="email", type="tel") |
| Consent | Checkbox only where legally required; pre-ticked boxes are not consent in the EU/UK |
Social proof hierarchy
| Type | Strength | Where |
| Specific testimonial with name, role, company, result | Very high | Near the first CTA and in the proof section |
| Star rating with count and platform | High | Hero, near CTAs |
| Recognisable logos | High (B2B) | Below hero |
| Case study link with headline number | High | Consideration pages |
| "[n] customers / [n] downloads" | Medium | Anywhere as a quick signal |
| Awards, certifications, security badges | Medium–high in regulated or payment contexts | Near form/payment |
Testimonial formula: "[Specific result in numbers or concrete change]" — [Name], [Role], [Company]. "Changed our business" is weak; "cut onboarding from 14 days to 2" is proof.
Ad-to-Page Message Match
The largest single lever for paid traffic. Score 1–5 on each; anything under 4 is leaking.
| Ad element | Page must match |
| Headline / hook | H1 uses the same words or the same promise |
| Offer (price, discount, guarantee, free trial) | Identical, visible above the fold |
| Audience named in the ad | Page names the same audience |
| Visual style and colour | Same palette, same imagery style |
| Tone | Same voice |
| Keyword (Search ads) | Appears in H1 or sub-headline |
Dynamic text replacement (by ad group or UTM) is the practical way to keep match high across many ads.
Speed and Mobile
Most paid traffic is mobile. Test on a real phone, not only a responsive preview.
| Metric | Target (Core Web Vitals, 75th percentile) |
| Largest Contentful Paint (LCP) | ≤ 2.5 s |
| Interaction to Next Paint (INP) | ≤ 200 ms |
| Cumulative Layout Shift (CLS) | ≤ 0.1 |
Usual fixes: compress and size the hero image, lazy-load below-the-fold media, remove unused scripts and heavy chat widgets, load the consent banner without blocking render, avoid layout shifts from late-loading fonts and banners.
Funnel Stage to Page Type
| Stage | Page type | Goal | CTA |
| Cold / awareness | Educational, guide, quiz | Capture email, qualify | Download, Take the quiz |
| Consideration | Feature overview, comparison, demo | Trial or call | Start trial, Book a demo |
| Decision | Offer / pricing | Purchase or sign-up | Buy, Sign up |
| Retargeting | Objection-handling, testimonial-heavy, offer | Overcome last barrier | Claim offer |
CRO Testing Framework
What to test first
| Element | Typical effect size | Why first |
| Offer (price framing, guarantee, trial length) | Largest | Changes the value equation |
| Headline / message match | Large | Decides whether people read on |
| Form length and steps | Large on lead gen | Direct friction |
| Hero visual (image vs product video) | Medium | Comprehension |
| Social proof placement and specificity | Medium | Belief |
| CTA label | Small–medium | Clarity |
| Colour and layout tweaks | Small | Test last |
Test design rules
- One hypothesis per test: "Adding a specific testimonial next to the form will raise form completion because trust is the blocker."
- Sample size: calculate it before launching (baseline rate, minimum detectable effect, 95% confidence, 80% power). Rule of thumb: a few hundred conversions per variant for a 10–15% lift; below ~100 conversions per variant results are mostly noise.
- Duration: full weeks (2+), never stop early on a peak, run through at least one business cycle.
- Hold constant: traffic sources, campaigns and offers between variants; split traffic randomly at the page level.
- Guardrails: watch lead quality and down-funnel conversion, not only the form rate.
- Tooling: Google Optimize is gone; use the testing tool the team has (VWO, Optimizely, AB Tasty, Convert, Unbounce/Instapage built-ins, or a server-side split). Low-traffic pages: run sequential, bolder changes and judge on direction plus qualitative feedback rather than p-values.
Reporting a test
HYPOTHESIS: [Change] will [effect] because [reason]
CONTROL: [n visitors] → [conversions] = [rate]
VARIANT: [n visitors] → [conversions] = [rate]
LIFT: [relative %] | CONFIDENCE: [%] | MDE planned: [%]
DOWN-FUNNEL CHECK: [lead quality / revenue per visitor]
DECISION: [Ship / iterate / discard] — [why]Output Format
LANDING PAGE AUDIT: [Page / URL]
Traffic source: [Ad / email / organic] | Goal: [conversion] | Current CVR: [x% or unknown]
Devices: [mobile % / desktop %] | Access: [fetched / from screenshots / from copy]
--- MESSAGE MATCH SCORE: X/5 ---
[Ad headline] ↔ [Page H1] — [gap and fix]
[Offer] ↔ [Page offer] — [gap and fix]
--- ABOVE THE FOLD ---
What / why / next: [pass or fail each, with the fix]
H1 rewrite: [..]
Sub-headline rewrite: [..]
CTA label: [current] → [recommended]
Risk reducer: [add ...]
--- BODY FLOW ---
[Current order] → [Recommended order]
Missing sections: [...]
Proof gaps: [`[insert proof]` placeholders and what to collect]
Objections unanswered: [...]
--- FORM ---
Fields: [current n] → [recommended n and which]
Steps / validation / label / privacy line / fallback: [changes]
--- SPEED & MOBILE ---
LCP / INP / CLS: [values or "not provided"] → [fixes]
Mobile issues: [...]
--- TEST PLAN (priority order) ---
1. [Hypothesis] — [metric] — [sample needed] — [duration]
2. [...]
3. [...]
--- QUICK WINS (ship without testing) ---
- [...]Common Pitfalls
- Several competing CTAs. Each extra action is a "no" to the main one. One primary action; repeat it; offer one lower-commitment fallback.
- Product-first copy. "Our AI-powered platform…" says nothing about the visitor. Lead with their outcome.
- Ignoring message match. A page that does not repeat the ad's promise makes visitors feel they clicked wrong. Match words, offer and visuals.
- Form before value. Show the benefit, then ask for details.
- Desktop-first design. Most paid traffic is mobile; a hero that needs two scrolls on a phone has no above-the-fold.
- Underpowered tests. Declaring a winner at 30 conversions per variant is a coin flip. Calculate sample size first.
- Optimizing vanity metrics. Time on page and bounce rate are not conversion. Judge on conversion rate, lead quality and revenue per visitor.
- Invented proof. Placeholder testimonials shipped to production are a legal and trust problem. Mark them and collect real ones.
Verification Checklist
- Message match scored against the actual ad/email that sends traffic
- Above the fold answers what / why me / what next on mobile without scrolling
- H1 and sub-headline rewritten with the visitor's words and a specific benefit
- One primary CTA, verb-first label, risk reducer beside it
- Body follows Problem → Proof → Solution → Steps → Social proof → Objections → CTA → FAQ
- Form fields reduced to what is needed now; validation, label, privacy line and fallback specified
- Social proof is specific (name, role, result) or marked
[insert proof] - Core Web Vitals targets stated with fixes where data exists
- Test plan has hypotheses, metrics, sample size and duration, in priority order
- Quick wins separated from things that need a test
- Page not described beyond what was actually fetched or provided
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

