Technical SEO & Core Web Vitals
Real request timing plus field and lab Core Web Vitals from Google PageSpeed Insights — never a guessed number. TTFB, LCP, CLS, INP, HTTPS, TLS, security headers and text compression, all measured.
SEO · 30%Everything measured · nothing estimated
VisibilityProof is a generative engine optimization tool that runs a deep technical and content audit across classic search and the AI answer engines in one pass — it finds the phrase your page actually targets, prices it against your market, checks where you rank and who the engines name instead of you, then hands you a ranked list of fixes written for your platform. Every figure is measured; nothing is modelled.
Four SEO pillars and four GEO pillars, weighted into one transparent score.
Real request timing plus field and lab Core Web Vitals from Google PageSpeed Insights — never a guessed number. TTFB, LCP, CLS, INP, HTTPS, TLS, security headers and text compression, all measured.
SEO · 30%The phrase your page is really built around — detected from your own title, headings, URL, meta and opening paragraph, not guessed from word counts — and where that phrase is missing. Plus title and meta length, heading order, a single H1, image alt coverage and byte weight, and internal and external link structure.
SEO · 30%Referring domains, domain rating and a toxic-link check from real backlink data — the off-page signals that still decide how you rank, measured rather than modelled.
SEO · 25%robots.txt and XML sitemap, canonical and noindex signals, HTTP status codes, redirect chains and a real 404 check — across a sampled crawl of your site, so you see what search engines actually reach.
SEO · 15%Whether ChatGPT, Perplexity, Gemini and Claude mention and cite your domain — asked using the phrases your page is actually built around, so the answer is about your subject rather than your most frequent word. The GEO surface classic SEO tools ignore entirely.
GEO · 35%We fetch your pages the way GPTBot and ClaudeBot do — without executing JavaScript — so we measure what an answer engine really sees. Critical content hidden behind JS is flagged, not assumed present.
GEO · 25%How liftable your content is for an AI answer — direct answers, question headings, definitions, structured data and llms.txt — the things that get a page quoted instead of skipped.
GEO · 20%Your share of the AI answers against the rivals competing for the same citations — who gets named when the engine can only pick a few.
GEO · 20%Paid plans add the deeper data sources — each measured live from its real API, gated to the plans that include it. Every one below is a real source, not a sample, and every figure arrives with the market it was priced against.
Six technical checks that need no extra request: whether your canonical points at this page or at a dead URL or another site, whether the indexing header and the meta tag contradict each other, what blocks the browser before it can draw anything, whether your structured-data blocks actually link to each other, broken pagination markup, and internal links that mix www and trailing-slash forms of the same page.
For every finding your audit raises, what to change and why it matters — written so you can act on it without guessing which part of the page it refers to.
The same fixes, written out for your exact CMS: which screen to open, which field to change, what to type, and how to check it worked afterwards. No technical knowledge assumed.
Referring domains, domain rating and a toxic-backlink percentage from real link data — not a modelled estimate.
Which domains the answer engines name instead of you, prompt by prompt, with your share of every naming event — and the sample size stated beside it.
Real monthly search volume and ranking difficulty for the terms your page actually targets, priced against your own market — measured, never modelled.
Repeat AI answer-engine checks over time, so you can watch your citation rate move as you ship fixes.
12-month Google search-interest trends and rising queries, so you write for demand that is climbing, not fading.
For local businesses: your Google Business Profile as customers see it — rating, reviews, category and opening hours, from the real listing. If your business has no profile, the card says so rather than scoring you against one.
How often the phrase your page targets is actually asked of AI answer engines — demand on the AI surface, not just in Google.
Where the page ranks in Google for the phrase it targets, the five results above it, and how their titles, headings and length compare with yours — measured against your own market.
What ChatGPT actually answers about your topic: the pages it cited, and the queries it expanded your question into on the way there.
Related phrases with real volume and ranking difficulty, and a verdict on the phrase your page targets today — advice about what to aim at, not another number to read.
VisibilityProof fingerprints the site the moment it scans — so its advice is grounded in what you actually run, not a generic checklist. Then it turns each finding into a fix, gated to paid plans.
Every issue is scored by severity × impact ÷ effort, so the top of the list is the highest-leverage change you can make today — not an alphabetical wall of warnings.
Premium data is served only after your entitlement is verified on the server — it is never sent to the browser and hidden with CSS. A free scan literally cannot leak paid data from the dev console.
Yes. VisibilityProof detects your CMS — WordPress, Shopify, Webflow, Wix, Squarespace or custom — and gives step-by-step fixes for that exact platform, including which plugin to use and where to click. Unknown CMS gets a clear, platform-agnostic explanation instead of wrong instructions.
Yes. It detects the CMS and version, the active theme, plugin slugs, framework, server and CDN — that part is an inventory, reported for reference rather than graded, because there is no right answer to which CMS you should be on. It also detects marketing and analytics trackers such as GA4, Google Ads, Meta Pixel, GTM and consent platforms, and those DO produce findings when something is wrong — an ad pixel firing before consent, or a tag installed twice.
The one your page is demonstrably built around. VisibilityProof reads your title, H1, subheadings, URL, meta description and opening paragraph and ranks the phrases that actually occupy them — it does not ask you to type a keyword, and it does not pick your most frequent word. It then tells you where that phrase is missing, and flags the case where your title declares one subject while your headings are built around another.
On Pro and above it takes your live Google position for the phrase the page targets, plus the five results above you and how their titles, headings and length compare with yours. On Agency it also reports which domains the answer engines name instead of you, prompt by prompt, with the sample size stated beside every share.
Every metric is measured live. Where a signal genuinely can’t be measured, VisibilityProof shows “not available” instead of inventing a figure — and it never models traffic or keyword volume the way some tools do. Search volume and difficulty are always shown with the market they were measured against, because the same phrase has different demand in different countries.
A free account audits any URL and shows scores, the report card, issue titles and keyword density, all measured live. The answer-engine card is a clearly labelled sample preview, not a live measurement — real queries to ChatGPT, Claude, Perplexity and Gemini cost money on every audit. Paid plans — each with a 14-day free trial — add those real queries plus the platform-specific fixes, backlink and keyword metrics, per-item lists, tracker fixes and CSV/PDF export.