How the Shazamme GEO Score works

The Shazamme Career Site Grader measures one thing classic SEO tools miss: whether your site is visible, credible and convertible across traditional search, AI answer engines (GEO/AEO) and the candidate journey. This page documents exactly how the score is produced — because a scoring standard should be transparent.

The three modes

The engine adapts its pillars and weighting to what is being graded:

• Recruitment — agencies and staffing firms (sector pages, apply flow, employer brand).
• Career Site — a single employer's careers/talent site (culture, EVP, DE&I, JobPosting schema).
• Website — any general site (security headers, UX, content quality).

Pillars & weighting

Each pillar is scored 0–100 from its own checks, then combined into the overall score using mode-specific weights. Recruitment mode, for example:

Within a pillar, every check carries its own point value; the pillar score is the points earned over points available, so adding checks automatically rebalances without distorting the weighting.

What makes this different: GEO & AEO

Generative AI is now an answer layer between candidates/clients and your site. We measure readiness for it directly:

• AI crawler access — whether GPTBot, ChatGPT Search (OAI-SearchBot), ClaudeBot, Google-Extended (Gemini/AI Overviews), PerplexityBot, Applebot-Extended and others are permitted in robots.txt.
• Entity & authority — Organization schema with sameAs links to Wikidata/LinkedIn/Crunchbase, consistent NAP, named people (E-E-A-T) and accreditations — how AI engines confirm who you are and whether to trust you.
• Structured-data validity — not just presence of schema but whether JobPosting carries the fields Google Jobs requires, plus a WebSite SearchAction.
• AEO / answer-engine readiness — summary/TL;DR blocks, question-style headings, tables and quotable stats: the content shape AI engines actually cite.
• Crawlable content — whether your content (including job listings) exists in the raw HTML or only after JavaScript renders, since most AI crawlers don't run JS.
• llms.txt / llm-info — emerging standards that tell AI models how to represent you.

Performance: real Core Web Vitals

Where available, performance uses Google PageSpeed Insights — real Chrome-user field data (CrUX) for LCP, INP and CLS, plus a Lighthouse lab score from a fully-rendered test — not a single server-timing guess.

Accuracy notes

• Multi-page & widget-aware: recruitment/career grades fetch the dedicated job routes (/job-results, /job-detail, etc.) and detect client-rendered job boards, so apply/search aren't falsely marked missing.
• Single-page snapshot: the deep content checks evaluate the submitted page plus those job routes — not a full-site crawl.
• Point-in-time: scores reflect the moment of the scan; sites change. Monitoring re-scores on a schedule to show the trend.

Grade scale