Analyze B2B SaaS homepages against 9 positioning checks from April Dunford and Fletch PMM. Full framework citations, PASS/FAIL verdicts, actionable fixes, and repositioning strategies.
claude /install positioning-grader@andorlabsOr load for a single session:
claude --plugin-dir /path/to/positioning-grader/positioning-grader:positioning <url>
One command does everything: scores all 9 checks, prints a scorecard to the terminal, and writes a full report with 3 repositioning strategies and homepage layout to ~/Desktop/claude-code/.
Based on April Dunford's Obviously Awesome (2019):
| Check | What it evaluates |
|---|---|
| DC-1 | Competitive alternatives acknowledged |
| DC-2 | Unique attributes surfaced |
| DC-3 | Value mapped to customer outcomes |
| DC-4 | Target customer identified |
| DC-5 | Market category clearly claimed |
Based on Fletch PMM's homepage teardown methodology (2020-2024):
| Check | What it evaluates |
|---|---|
| FQ-1 | "What is it?" answered in 5 seconds |
| FQ-2 | "Who is it for?" explicitly stated |
| FQ-3 | "What does it replace?" made clear |
| FQ-4 | "Why is it better?" demonstrated |
Every check produces:
- What it is — from the framework (word for word)
- Why it matters — the positioning rationale
- Source — full citation (author, book/methodology, year)
- Compliance — PASS / FAIL / N/A
- Observation — what was found on the homepage
- How to fix — actionable recommendation (FAIL only)
Only the homepage is analyzed — positioning is a homepage concern.
- Claude Code CLI
- WebFetch (built-in)
- Chrome (optional, for homepage screenshot)
MIT