Landing copy

Build the page around a defensible buying claim and the evidence needed to evaluate it. Preserve an effective brand voice; do not force every product into terse, casual, benefit-first startup copy.

Choose the mode and page job

• Positioning — find the claim, audience, alternatives, mechanism, and proof before page copy.
• Audit — diagnose existing messaging and propose prioritized repairs.
• Targeted rewrite — improve a hero, proof section, pricing explanation, objection, or CTA in context.
• Generate — produce a complete page message architecture and copy deck.
• Experiment — create variants that test one meaningful messaging hypothesis.

Identify the page job: category introduction, product selection, campaign response, feature evaluation, waitlist signup, pricing decision, or another explicit action. Determine traffic intent, buyer awareness, purchase risk, category maturity, and available proof from the brief, repository, customer language, analytics, or supplied research.

Ask only for missing facts that would change the claim, audience, offer, or legal truth. When facts remain unavailable, mark them as content requirements rather than inventing them.

Find the defensible claim

Read claim-first writing before positioning, generate mode, or any rewrite that changes the core proposition.

Document:

• buyer and situation;
• failure, cost, or unmet goal;
• product capability and mechanism;
• close alternatives and meaningful difference;
• evidence, constraint, demonstration, or costly promise;
• exact customer or domain language;
• limits the page must not imply away.

Generate several plain claim shapes, then choose or recommend the one best supported by importance, specificity, differentiation, and proof. Confirm with the user only when choosing among materially different business promises; otherwise proceed with stated assumptions.

Audit the decision path

Evaluate the page as a sequence rather than scoring isolated ingredients:

1. Can the intended buyer recognize the page is for their situation?
2. Is the proposition understandable and important?
3. Does the page explain or demonstrate how the product changes the outcome?
4. Is differentiation visible where alternatives enter the decision?
5. Does proof support the exact claim being made?
6. Are price, trust, risk, compatibility, effort, and limitations handled near the decisions they affect?
7. Does each CTA accurately describe the next step and commitment?
8. Does the voice match the brand, audience, and stakes?

Report missing evidence separately from writing defects. Do not lower copy quality merely because the company has not supplied proof; identify the content gap and narrow unsupported claims.

Design the message architecture

Select only the sections required for this buying decision. Possible roles include:

• proposition and next action;
• product evidence or demonstration;
• mechanism and capabilities;
• comparison or alternatives;
• proof and provenance;
• use-case qualification;
• pricing context and terms;
• objections, limitations, security, or trust;
• final decision support.

Section order follows reader questions and evidence, not a universal hero/features/testimonials/FAQ template. A familiar product with high-intent traffic may need little education; a new category or high-risk purchase may require mechanism, comparison, and proof before action.

Write without flattening the voice

Use concrete domain nouns, actions, mechanisms, constraints, and observable results. Keep technical detail when the buyer needs it. Use customer language when it is representative and accurately sourced.

Headline length, sentence length, feature-first versus benefit-first order, casing, contractions, fragments, humor, and punctuation are contextual choices. Reject filler and unsupported superlatives because they add no substance, not because individual words are universally banned.

Calls to action should make destination, result, or commitment clear in context. Conventional labels can be appropriate when surrounding UI makes them unambiguous. Do not manufacture urgency, scarcity, customer counts, rankings, or performance outcomes.

Read copy patterns and examples only when selecting a narrative framework, diagnosing a recurring copy failure, or designing an experiment. Treat patterns as options, not performance guarantees.

Design experiments honestly

Test one material hypothesis at a time, such as audience framing, problem versus outcome emphasis, mechanism visibility, proof placement, or commitment language. Keep offer, layout, and traffic stable when the goal is to attribute messaging impact.

State the hypothesis, primary measure, guardrail measure, audience, sample or runtime limitation, and decision rule. Do not predict arbitrary lift percentages.

Output contract

Audit or targeted rewrite

Page job and intended buyer:
Core claim and proof status:
Priority findings: [impact, evidence, and source location]
Before -> after: [for grounded repairs]
Content required: [missing facts, proof, or product decisions]
Preserve: [effective language or structure]
Verification and experiment recommendations:

Generate or positioning

Deliver:

1. claim/evidence ledger;
2. message architecture with the question each section answers;
3. complete copy deck labelled by section and component;
4. CTA destinations and commitment level;
5. proof, asset, and content requirements;
6. assumptions, excluded claims, and implementation notes;
7. optional variants only when they test a stated hypothesis.

Verify

1. Map every material claim to product behavior, a source, a scoped assumption, or an explicit content requirement.
2. Run a competitor/alternative swap test on major propositions and explain any generic line retained.
3. Trace the reader's likely questions through the section order; remove repeated promises that add no mechanism, proof, or decision support.
4. Check price, security, privacy, compatibility, cancellation, implementation effort, and limitations where relevant to the offer.
5. Confirm CTA text, destination, and actual commitment agree.
6. Read the page aloud and compare it with established brand/customer language without normalizing every sentence into one cadence.
7. Check terminology consistency across landing page, product UI, pricing, and documentation.
8. Verify no proof, statistic, urgency, guarantee, or comparison was invented or broadened.
9. For implemented copy, inspect responsive rendering, text overflow, links, analytics hooks, and accessible control names.
10. Report what was verified directly and what still needs customer, legal, product, or analytics input.