B2B AI DirectoryB2B AI Directory
Website ManagementintermediatePro

Build a homepage positioning audit skill from your own frameworks

Turn your positioning framework, saved examples, and homepage review standards into a repeatable Claude Code skill that grades B2B homepages and produces concrete copy fixes.

What you will have

A reusable homepage audit skill that reviews any B2B website URL, scores the hero and page, diagnoses positioning gaps, and outputs rewrite-ready recommendations.

Setup time
3-5 hours for the first usable skill
Time saved
3-5 hours per homepage review
Estimated cost
$20 to $200 per month
Tools used
5 tools

Why this works

Most homepage feedback is too subjective: one person reacts to the hero, another argues about visual design, and nobody uses the same standard twice. A skill forces the reviewer to apply the same positioning questions, scoring rules, and examples every time. The real value is not that Claude writes better headlines on the first pass; it is that your review process becomes explicit, testable, and easier to improve after every audit.

Step-by-step workflow

Preview the workflow

The first 2 steps are open. Pro unlocks the remaining steps, copy-paste prompts, pro tips, tool-by-tool setup guidance, and implementation details.

1

Define the audit scope, audience, and decisions it must support

45-60 min

Define whether the skill will evaluate only the homepage, the homepage plus linked product pages, or a specific page section such as the hero, proof, CTA, or category explanation. In Google Sheets, create fields for Audit Audience, Company Stage, Category Maturity, Target Buyer, Primary Conversion Goal, Page Scope, Decisions Supported, Required Output, Review Owner, and Acceptance Criteria. State whether the audit is intended to support a copy refresh, positioning review, redesign brief, launch decision, competitive benchmark, or agency deliverable. Separate strategic questions that require research from copy issues the skill can safely identify from the page and approved sources. Define what the skill must not judge, such as visual design quality, SEO, accessibility, technical performance, or product-market fit unless those criteria are explicitly included. Set the expected audit length, completion time, evidence standard, and level of confidence required before recommending a rewrite. Review the scope with Product Marketing, Website Marketing, and the person who will implement the findings.

Output

An approved audit charter defining scope, audience, supported decisions, exclusions, evidence standard, deliverables, and acceptance criteria.

Google SheetsGoogle Docs
Pro tip

A homepage skill becomes inconsistent when it is quietly asked to evaluate positioning, visual design, conversion, SEO, and brand at the same time.

2

Curate the framework, example, and reference library

90-120 min

Create a Google Drive folder named `homepage-positioning-skill-sources` with subfolders for Frameworks, Strong Examples, Weak Examples, Stage Adjustments, Category Notes, CTA Rules, Proof Rules, Rewrite Style, and Archived Sources. Copy the relevant sections from internal documents, Google Docs, and Substack posts into stable reference files rather than relying on links that may change or require access later. Include at least three strong and three weak homepage examples with notes explaining why each example belongs in that category. Remove duplicate, obsolete, contradictory, or unsupported guidance and mark any source that represents opinion rather than an approved operating rule. Record author, source date, intended audience, company stage, category context, and usage rights for every file. Avoid placing confidential customer, employee, or unpublished company information into a broadly reusable skill source folder. Ask the framework owner to approve the initial library before the skill is built.

Output

A reviewed source library containing approved frameworks, annotated examples, stage and category guidance, CTA and proof rules, rewrite style, and source metadata.

Google DriveGoogle DocsSubstack
Pro tip

Examples need commentary. Without an explanation of why a page is strong or weak, the skill may learn superficial patterns instead of the framework.

Pro workflow preview

Previewing 2 of 16 steps

Pro membership

Unlock the full workflow

Get the remaining 14 steps, copy-paste prompts, pro tips, tool-by-tool setup guidance, and weekly new workflows.

$9/month

Create the source index and evidence hierarchy
Build the human benchmark set and expected audit outcomes
Initialize the Claude Code workspace and operating instructions
Define the audit output contract before writing the skill
Define the scoring rubric and calibration rules
Build the first complete Markdown skill and supporting templates
See Pro plan
3Create the source index and evidence hierarchy
Locked
4Build the human benchmark set and expected audit outcomes
Locked
5Initialize the Claude Code workspace and operating instructions
Locked
6Define the audit output contract before writing the skill
Locked
7Define the scoring rubric and calibration rules
Locked
8Build the first complete Markdown skill and supporting templates
Locked
9Run a controlled audit on one known homepage
Locked
10Benchmark the skill across the full calibration set
Locked
11Review disagreements and classify repeatable versus one-off defects
Locked
12Update the skill only from approved repeatable findings
Locked
13Run regression tests and approve the production skill version
Locked
14Run a production audit and create the implementation backlog
Locked
15Validate recommendations before copy is published
Locked
16Operate the monthly benchmark and governed skill-improvement loop
Locked

Expected results

Audit consistency

Reusable skill file

The same rubric, source hierarchy, and output format are applied across homepages instead of relying on ad hoc feedback.

Review time saved

3-5 hours per audit

The skill handles structured diagnosis and first-pass rewrites while humans review judgment, priorities, and strategic tradeoffs.

Implementation output

Prioritized copy backlog

The workflow ends with section-level fixes, owners, and confidence notes instead of a long memo that nobody acts on.

Skill maturity

Improves after every real example

Calibration logs and version notes let the skill compound from actual homepage tests rather than staying as a generic prompt.

Related workflows

Continue with workflows that share a similar GTM motion, category, or tool stack.