Core Web Vitals with real field data, indexation directives, structured-data coverage, crawlability and AI-crawler access — exact values, line-item fixes, raw JSON export.
Built for developers & marketers · free account · 3 audits a month
We crawl your main pages the way a search engine does — plain HTML first, headless rendering when your content needs JavaScript — then run every check below and report the raw values.
LCP, INP and CLS from real-user CrUX field data where Google has it, plus Lighthouse lab runs for mobile and desktop — exact values, not just colors.
Robots meta directives (noindex/nofollow), meta refresh, canonical tag presence, 404 handling and www-canonicalization — the silent index killers.
JSON-LD detection across your main pages: which schema types are present (Article, FAQ, HowTo, Organization…) and which pages have none.
robots.txt rules, sitemap.xml presence and URL counts, ads.txt — whether crawlers can discover and are allowed to fetch what matters.
Internal/external link balance, follow vs nofollow distribution, plaintext emails and unsafe cross-origin links across your crawled pages.
Viewport configuration and responsive-design signals — the prerequisites for mobile-first indexing.
HTTPS, HSTS, mixed content, HTML compression, SPF/DMARC DNS records and server/CDN fingerprinting.
robots.txt access for GPTBot, ClaudeBot, PerplexityBot, Google-Extended, CCBot and Applebot-Extended, plus llms.txt — can AI answer engines even read you?
Your report shows where each metric lands — from real-user field data when Google has enough traffic for your site, and lab runs either way.
| Metric | Good | Needs improvement | Poor |
|---|---|---|---|
| LCP — Largest Contentful Paint | ≤ 2.5s | 2.5s – 4.0s | > 4.0s |
| INP — Interaction to Next Paint | ≤ 200ms | 200ms – 500ms | > 500ms |
| CLS — Cumulative Layout Shift | ≤ 0.1 | 0.1 – 0.25 | > 0.25 |
A stray noindex, a JavaScript-only page, a canonical pointing the wrong way — technical failures don't degrade rankings gradually, they remove pages from the game entirely. And the newest gap is invisible in classic tools: if your robots.txt blocks AI crawlers, you don't exist in ChatGPT or Perplexity answers either. This audit surfaces both, with the failing page and directive named.
Every check is deterministic and rule-based: we fetch your pages (rendering JavaScript when needed), evaluate each signal against a fixed rubric, and roll results up into five category grades — On-Page, Links, Usability, Performance and Social — plus an AI Search Readiness score. Performance uses Google's PageSpeed Insights and CrUX data directly. The AI layer only ranks and explains what the rules found; it never invents findings. Same input, same score — which is what makes re-runs comparable over time.
Exact metric values, per-check pass/fail with the failing element named, and raw JSON export — a bug report, not a sales pitch.
The technical evidence to get engineering time prioritized — every finding carries a plain-English business impact you can put in a ticket.
Audit a prospect or client site in a minute, share the read-only report link, and export the raw data into your own tooling.
Core Web Vitals (LCP, INP, CLS) are Google's official metrics for load speed, responsiveness and visual stability, and they factor into ranking. The audit reports both lab measurements and real-user (CrUX field) data where Google has it, with the exact values — not just a color.
Yes. The audit checks robots meta directives (noindex/nofollow), robots.txt rules, sitemap.xml presence and URL counts, canonical tag presence, meta refresh redirects, and how your site handles 404s — the usual suspects when pages silently drop out of the index.
It detects which JSON-LD types are present on your pages (Article, FAQ, HowTo, Organization and more) and flags pages missing structured data entirely. It reports presence and type coverage rather than validating every property against the schema.org spec.
We test whether GPTBot, ClaudeBot, PerplexityBot, Google-Extended, CCBot and Applebot-Extended are allowed or blocked by your robots.txt, and whether you publish an llms.txt. If AI assistants can't crawl you, they can't cite you — this is the newest visibility gap most audits ignore.
This version of the report leads with raw metrics and exact values for a technical audience. Every finding still carries a plain-English explanation, so it works as a shared artifact between marketing and engineering.
Yes — every report can be downloaded as raw JSON (the complete result object, every check and metric), alongside the print-friendly PDF view and a shareable read-only link.
Exact metrics, named failures, raw JSON export. If the findings need engineering muscle, our team builds too.
Talk to our teamCompare your site against a competitor · Focus on AI search readiness · Prefer the simpler version?