gtm-grader

GTM Grader

Score any SaaS website against 77 research-backed checks covering SaaS optimization (47 rules from peer-reviewed marketing science), homepage positioning (9 checks from Dunford and Fletch PMM), and search discoverability (21 checks from Google documentation, SEO research, and content marketing studies). Every check includes its full source citation.

Commands

Command	What it does	Output
/gtm-grader:grade <url>	47-rule SaaS optimization analysis	[company]-saas-analysis.md
/gtm-grader:position <url>	9-check positioning diagnostic	[company]-positioning-analysis.md
/gtm-grader:search <url>	21-check search gap analysis + competitive landscape + content pillars	[company]-search-gap.md
/gtm-grader:wins <url>	One-page prioritized summary (both)	[company]-gtm-wins.md
/gtm-grader:export <url>	Full combined analysis (both)	[company]-gtm-export.md

Installation

Add the plugin directory to your Claude Code settings file (~/.claude/settings.json):

{
  "permissions": {
    "allow": [
      "mcp",
      "Bash(grep:*)",
      "Read",
      "Write"
    ]
  },
  "plugins": [
    "/path/to/gtm-grader"
  ]
}

Or load it for a single session:

claude --plugin /path/to/gtm-grader

Requirements

• Claude Code CLI
• WebFetch capability (built into Claude Code)
• Google Chrome (for homepage screenshots — optional, text-only analysis works without it)

What It Checks

SaaS Optimization (47 Rules)

#	Category	Rules	Examples
1	Brand & Messaging	BM-1 to BM-8	Productized headline, top 3 benefits, present tense, newness, simple language, social proof numbers, star rating, curated testimonial
2	Page Design & Visuals	PD-1 to PD-14	Design-product fit, brand colors, layout adaptation, complexity, rounded CTAs, machine font, CTA placement, before/after flow, video usage, above-the-fold, succinctness, scannable structure, signal-to-noise
3	Pricing Plans	PR-1 to PR-9	3-5 plans, decoy plan, left-to-right ordering, price below features, difference framing, simple pricing, easy math, flat+usage hybrid, flat-rate option
4	Free Trials	FT-1 to FT-4	Full-feature trial, 7-day length, high trial usage, extensions over discounts
5	Freemium	FM-1 to FM-2	Usage-limited freemium, freemium decoy plan
6	Churn Prevention	CP-1 to CP-3	Customer gifts, auto-enrolling loyalty, generous downgrades
7	Affiliates	AF-1 to AF-2	Optimize affiliate payment model, protect from deceptive affiliates
8	Referrals	RF-1 to RF-5	Altruistic framing, pre-filled messages, avoid money rewards, keep rewards small, targeted referrals

Sources: Journal of Marketing, Journal of Consumer Research, Marketing Science, Management Science, and 10+ other peer-reviewed journals. Research from UCLA, Stanford, MIT Sloan, Princeton, NUS, Nielsen Norman Group, and others.

Search Discoverability (21 Checks)

#	Category	Checks	Examples
9	Technical SEO	TS-1 to TS-8	Schema.org structured data, meta titles, meta descriptions, heading hierarchy, internal linking, image optimization, canonical tags, OG/Twitter cards
10	E-E-A-T Signals	EA-1 to EA-7	Author attribution, About page depth, original research, external citations, trust signals, case studies, content freshness
11	Content Depth	CD-1 to CD-6	Content depth per page, topic coverage, FAQ content, visual quality, pillar/cluster architecture, funnel balance

Plus: competitive landscape analysis (top 3 SERP competitors) and scored content pillar recommendations.

Sources: Google Search Central, Google Quality Rater Guidelines, web.dev, Moz, Ahrefs, Backlinko, SEMrush, HubSpot, Content Marketing Institute.

Positioning (9 Checks)

#	Category	Checks	Framework
12	Dunford Canvas	DC-1 to DC-5	Competitive alternatives, unique attributes, value mapping, target customer, market category
13	Fletch Questions	FQ-1 to FQ-4	What is it?, Who is it for?, What does it replace?, Why is it better?

Sources: Dunford, A. Obviously Awesome (2019). Pierri, A. & Fralic, R. Fletch PMM (2020-2024).

Priority Classification (wins command)

The /wins command classifies failures into three tiers:

• Critical — Above-the-fold items that directly block conversion: headline misses, missing social proof, broken CTA placement, unclear value proposition, unidentified target customer
• Medium — Below-the-fold or moderate-impact items: design mismatches, pricing structure issues, content organization problems
• Low — Less visible or smaller effect size items: font choices, video speed, referral details, affiliate structure, churn signals

Extensibility