Scoring and pain grades
How ten category scores become one 0-100 composite, and why a lower score means a hotter lead.
Ten category scores, one weighted composite, one letter grade. The scale runs backwards from every other score you have seen, and that inversion is the product.
Lower is hotter
A composite of 100 describes a fast, secure, accessible, well-marked-up website. It is also a business with nothing wrong to sell against. A composite of 30 describes a business losing customers to problems you can fix this month and prove you fixed. The engine ranks your queue by pain, not by quality.
The ten categories
Category | Default weight | What feeds it |
|---|---|---|
Performance |
| Core Web Vitals, lab performance, TTFB, page weight |
Security |
| TLS grade, security headers, known CVEs, exposed surface |
SEO |
| Indexability, metadata, canonicalisation, structured data |
Reputation |
| Review volume, rating, recency, owner responses |
Conversion |
| Calls to action, contact paths, forms, trust signals |
Accessibility |
| WCAG failures found on the rendered page |
Tech stack |
| Framework currency, CMS health, dependency risk |
AI visibility |
| AI-crawler access, answerable content, citation signals |
Authority (E-E-A-T) |
| Entity definition, expertise and trust signals |
Content quality |
| Substance, freshness, placeholder content |
Grades
Grade | Composite | Read as |
|---|---|---|
A |
| Healthy; little to sell |
B |
| Healthy; little to sell |
C |
| Worth a look; a single strong angle may exist |
D |
| Hot lead; significant, provable problems |
F |
| Hot lead; significant, provable problems |
Missing signals score neutral
If a category has no usable measurement, because the accessibility scanner was not installed, say, it is scored neutral rather than zero. Scoring an unmeasured category as a failure would manufacture pain that is not there, and the first prospect to notice would be right to dismiss the whole report.
Weighting for what you actually sell
The same city should not rank identically for a security consultancy and an SEO shop. Scoring profiles re-weight the categories around what you fix, and the profile that produced a score is stored with it so an old score is never silently reinterpreted under new weights.
default: Balanced (all services)seo_agency: SEO / content agencyweb_design: Web design / build agencysecurity: Security / IT servicesreputation: Reputation / local SEO
From score to pitch
Scoring also selects the outreach angles: the findings that are severe, provable and sellable, ranked. That ranked list is what the report's action plan opens with and what the outreach draft is built around, so the number, the report and the email always tell one consistent story.