What the RELACON website check examines

The check runs 193 rules in four areas. Every rule below explains what is measured, why it matters, how a fix looks and where the requirement comes from.

Four areas

SEO basics: HTTPS, redirects, indexability, title, description, canonical, headings, links, URLs, Open Graph, robots.txt, sitemap, 404.
AI visibility (GEO): AI crawlers, snippet control, entity, answer-ready content, llms.txt.
Structured data: JSON-LD validity, WebSite, Organization, Person, breadcrumbs, FAQ, articles, ratings, events, videos.
Technology and security: mobile, security headers, caching, images, load signals, icons and manifest, accessibility, analytics consent, email domain (SPF, DKIM, DMARC, MTA-STS), security.txt, bot access.

What is fetched

The page itself, its http and www variants, robots.txt, sitemap.xml, llms.txt, security.txt, favicon, web app manifest, a non-existent URL for the 404 test, the MTA-STS policy, up to five images and the og:image. The page is also fetched once as Googlebot and once as GPTBot (our probe names itself) to find bot challenges. DNS records are read over DNS-over-HTTPS.

SEO basics (64 checks)

HTTPS

SeverePage loads over HTTPSseo.https.loads

What is checked: Whether the page is reachable and ends up being served over HTTPS.
Why it matters: Browsers flag HTTP pages as not secure and Google prefers HTTPS. Without encryption, third parties can read or alter the content.
Example: `https://www.example.com/` instead of `http://www.example.com/`
Source: web.dev: Why HTTPS matters

SevereHTTP redirects to HTTPS with a 301seo.https.redirect

What is checked: Whether the `http://` address of the page permanently (301 or 308) redirects to the HTTPS address.
Why it matters: Old links and typed addresses without `https` would otherwise land on the unencrypted version. A permanent redirect gives search engines one clear main URL.
Example: `http://example.com/` → `301` → `https://example.com/`
Source: Google Search Central: Redirects and Google Search

SevereValid TLS certificateseo.https.cert

What is checked: Whether the HTTPS connection can be established without certificate errors.
Why it matters: With an expired or mismatched certificate, browsers show a warning page and crawlers abort the request.
Example: Certificate for `example.com` and `www.example.com`, renewed automatically (e.g. Let's Encrypt).
Source: MDN: Transport Layer Security

MediumHSTS setseo.https.hsts

What is checked: Whether the response includes the `Strict-Transport-Security` header.
Why it matters: With HSTS, browsers only load the site over HTTPS, even when someone types `http://`. This removes the first unencrypted request.
Example: `Strict-Transport-Security: max-age=63072000; includeSubDomains`
Source: MDN: Strict-Transport-Security

Redirects

Mediumwww and non-www consistentseo.redirect.www

What is checked: Whether the variant with or without `www` ends up on the same host as the checked page.
Why it matters: If both variants serve content, search engines see duplicate pages and signals are split across two addresses.
Example: `https://www.example.com/` → `301` → `https://example.com/`
Source: Google Search Central: Consolidate duplicate URLs

MediumAt most 2 redirect hopsseo.redirect.hops

What is checked: How many redirects it takes until the final page answers.
Why it matters: Every hop adds load time, and crawlers do not follow long chains indefinitely.
Example: `http://example.com` → `https://example.com/` in one step instead of via `www` and a language path.
Source: Google Search Central: Redirects and Google Search

LowNo 302 for permanent redirectsseo.redirect.302

What is checked: Whether a temporary redirect (302 or 307) occurs on the way to the page.
Why it matters: A temporary redirect tells search engines the old address will come back. For permanent moves, 301 or 308 is the clearer signal.
Example: `HTTP/1.1 301 Moved Permanently` instead of `302 Found`
Source: Google Search Central: Redirects and Google Search

Indexability

SevereStatus 200seo.index.status

What is checked: Whether the page answers with HTTP status 200.
Why it matters: Only pages with status 200 are indexed normally. Error codes cause the page to drop out of the index.
Example: `HTTP/2 200`
Source: Google Search Central: How HTTP status codes affect Google Search

SevereNo noindexseo.index.noindex

What is checked: Whether the robots meta tag or the `X-Robots-Tag` header contains `noindex`.
Why it matters: With `noindex` the page does not appear in search results. It is often left over from a staging site after a relaunch.
Example: `<meta name="robots" content="index, follow">` or leave the tag out entirely.
Source: Google Search Central: Block indexing with noindex

SevereNot blocked by robots.txtseo.index.robotsblocked

What is checked: Whether a `Disallow` rule in robots.txt blocks the page's path for Googlebot.
Why it matters: Google cannot read blocked pages. They appear without a description at best, or not at all.
Example: `User-agent: *` with `Disallow: /internal/` instead of `Disallow: /`
Source: Google Search Central: Introduction to robots.txt

Title

SevereTitle presentseo.title.present

What is checked: Whether the page has a non-empty `<title>` element.
Why it matters: The title is the basis for the headline in search results and the name in the browser tab.
Example: `<title>Tax advice in Graz | Example Ltd</title>`
Source: Google Search Central: Influencing your title links

MediumTitle 30–60 charactersseo.title.length

What is checked: Whether the title is between 30 and 60 characters long.
Why it matters: Very short titles say little, very long ones are cut off in search results. Google sets no fixed limit; the range is a guideline.
Example: `Tax advice in Graz | Example Ltd` (32 characters)
Source: Google Search Central: Influencing your title links

MediumOnly one title elementseo.title.unique

What is checked: Whether the HTML contains exactly one `<title>` element.
Why it matters: With several titles it is unclear which one applies. This often happens when both a theme and a plugin add one.
Example: A single `<title>` in the `<head>`.
Source: MDN: <title> element

LowTitle not identical to H1seo.title.h1

What is checked: Whether the title and the first H1 are word for word the same.
Why it matters: Title and H1 do different jobs: the title sells the page in search results, the H1 guides the reader on the page. Two different phrasings make use of both.
Example: Title `Tax advice in Graz | Example Ltd`, H1 `Your tax adviser in Graz`
Source: Google Search Central: Influencing your title links

MediumTitle has at least two wordsseo.title.words

What is checked: The title consists of at least two words.
Why it matters: A single word tells search engines little about the page.
Example: Wrong: “Home”. Right: “Web design for small businesses”.
Source: Google Search Central

LowTitle names more than the brandseo.title.notbrand

What is checked: The title contains a topic in addition to the brand name.
Why it matters: A title with only the company name is rarely clicked in search results.
Example: Wrong: “Muster Ltd”. Right: “Tax advice Graz | Muster”.
Source: Google Search Central

Meta description

MediumMeta description presentseo.meta.present

What is checked: Whether the page has a `<meta name="description">` with content.
Why it matters: Google often uses the description as the text below the search result. Without it, Google picks an excerpt itself, which does not always fit.
Example: `<meta name="description" content="Tax advice for freelancers in Graz. Free first consultation.">`
Source: Google Search Central: Control your snippets in search results

LowMeta description 70–160 charactersseo.meta.length

What is checked: Whether the meta description is between 70 and 160 characters long.
Why it matters: Short descriptions waste space in the search result, long ones are cut off. The range is a guideline, not a Google requirement.
Example: Two short sentences: offer, location, next step.
Source: Google Search Central: Control your snippets in search results

MediumMeta description is not a placeholderseo.meta.placeholder

What is checked: The meta description contains original text instead of a placeholder or CMS default.
Why it matters: Placeholders appear verbatim in search results and look untrustworthy.
Example: Wrong: “Just another WordPress site”. Right: a sentence about the page.
Source: Google Search Central

LowMeta description differs from the titleseo.meta.nottitle

What is checked: The meta description and the title are not the same text.
Why it matters: Two identical texts waste space in the search result.
Example: Wrong: title “Muster | Web design” and description “Muster - Web design”.
Source: Google Search Central

Canonical

MediumCanonical presentseo.canonical.present

What is checked: Whether the page has a `<link rel="canonical">` with an address.
Why it matters: The same page is often reachable under several addresses (with parameters, with a trailing slash). The canonical says which address belongs in the index.
Example: `<link rel="canonical" href="https://example.com/services/">`
Source: Google Search Central: Consolidate duplicate URLs

LowCanonical is absoluteseo.canonical.absolute

What is checked: Whether the canonical is a full address with `https://` and the domain.
Why it matters: Relative values usually work but easily point to the wrong address with copies, proxies or a wrong base URL. Google recommends absolute URLs.
Example: `https://example.com/services/` instead of `/services/`
Source: Google Search Central: Consolidate duplicate URLs

MediumCanonical points to a reachable pageseo.canonical.self

What is checked: Whether the canonical points to the same domain as the checked page.
Why it matters: A canonical pointing to another domain (such as the old domain or a staging site) moves the page in the index over there.
Example: Page `https://example.com/blog/` with canonical `https://example.com/blog/`
Source: Google Search Central: Consolidate duplicate URLs

MediumCanonical matches the served URLseo.canonical.matches

What is checked: The canonical names the same address the server serves, with trailing slash, letter case and default port normalised.
Why it matters: If the canonical points elsewhere, Google may index the wrong address.
Example: Served: `/contact` → `<link rel="canonical" href="/contact">`.
Source: Google Search Central

LowCanonical without tracking parametersseo.canonical.noquery

What is checked: The canonical contains no tracking or session parameters; pagination is allowed.
Why it matters: Parameters create duplicates of the same page.
Example: Wrong: `?utm_source=newsletter`. Right: `/product`.
Source: Google Search Central

Headings

MediumExactly one H1seo.headings.h1

What is checked: How many `<h1>` headings the page has.
Why it matters: One H1 names the main topic of the page. If it is missing or there are several, it is less clear to readers, screen readers and search engines what the page is about.
Example: `<h1>Tax advice in Graz</h1>`, followed by `<h2>` sections
Source: MDN: <h1>–<h6> elements

LowHeading order without skipped levelsseo.headings.order

What is checked: Whether heading levels follow each other without skipping (e.g. not from H2 straight to H4).
Why it matters: Screen reader users navigate a page by its headings. An outline without gaps helps them and makes the structure unambiguous for machines.
Example: H1 → H2 → H3 → H2
Source: MDN: <h1>–<h6> elements

LowSubheadings (H2) presentseo.headings.h2

What is checked: The page structures its content with at least one H2.
Why it matters: Subheadings help readers, screen readers and search engines grasp the content.
Example: `<h2>Services</h2>` below the H1.
Source: Google Search Central

Language and hreflang

Mediumlang attribute on the html elementseo.lang

What is checked: Whether the `<html>` element has a `lang` attribute.
Why it matters: The language setting drives pronunciation in screen readers, hyphenation and automatic translation.
Example: `<html lang="en">`
Source: MDN: lang attribute

Lowhreflang valid (with x-default)seo.hreflang

What is checked: Whether existing `hreflang` entries use valid language codes and include an `x-default`. Single-language sites need none.
Why it matters: With hreflang, Google shows each user the matching language version. Invalid codes are ignored; `x-default` covers all other languages.
Example: `<link rel="alternate" hreflang="en-GB" href="https://example.com/">` and `hreflang="x-default"`
Source: Google Search Central: Localized versions of your pages

Open Graph

Mediumog:title presentseo.og.title

What is checked: Whether the page has a `<meta property="og:title">` with content.
Why it matters: LinkedIn, Facebook, Slack and messengers show og:title as the headline of the link preview.
Example: `<meta property="og:title" content="Tax advice in Graz">`
Source: The Open Graph protocol

Mediumog:description presentseo.og.description

What is checked: Whether the page has a `<meta property="og:description">` with content.
Why it matters: The description appears below the headline of the link preview and helps decide whether someone clicks.
Example: `<meta property="og:description" content="Free first consultation, appointments online too.">`
Source: The Open Graph protocol

Mediumog:image reachable and at least 1200 px wideseo.og.image

What is checked: Whether an og:image is set, whether it can be loaded and whether it is at least 1200 px wide.
Why it matters: Without an image, a shared page looks plain. Images that are too small are shown as a small tile instead of a large preview.
Example: `<meta property="og:image" content="https://example.com/og.jpg">` at 1200×630 px
Source: The Open Graph protocol

Lowog:type presentseo.og.type

What is checked: The page declares its content type with `og:type`.
Why it matters: Social networks present the content more appropriately.
Example: `<meta property="og:type" content="website">`
Source: Open Graph protocol

Mediumog:image is a full HTTPS URLseo.og.imageabsolute

What is checked: The preview image is given as a full HTTPS URL.
Why it matters: Relative addresses and http fail to load a preview on some services.
Example: Right: `https://example.com/og.jpg`.
Source: Open Graph protocol

Lowog:url matches the canonicalseo.og.url

What is checked: `og:url` names the same address as the canonical, otherwise the served page.
Why it matters: Mismatched addresses create duplicates when sharing.
Example: `<meta property="og:url" content="https://example.com/page">`
Source: Open Graph protocol

Lowog:image with width and heightseo.og.imagesize

What is checked: The preview image width and height are declared in the head.
Why it matters: Without dimensions the preview loads later and is sometimes cropped wrongly.
Example: `og:image:width` 1200 and `og:image:height` 630.
Source: Open Graph protocol

Lowog:locale in the de_AT formatseo.og.locale

What is checked: The page language is declared in `og:locale` as language_COUNTRY.
Why it matters: Platforms then show the preview in the right language.
Example: `<meta property="og:locale" content="de_AT">`
Source: Open Graph protocol

X cards (Twitter)

Lowtwitter:card presentseo.og.twitter

What is checked: Whether the page has a `<meta name="twitter:card">`.
Why it matters: X (Twitter) uses twitter:card to pick the preview format. With `summary_large_image` the image is shown large.
Example: `<meta name="twitter:card" content="summary_large_image">`
Source: X Developer Platform: Cards markup

Lowtwitter:card is summary_large_imageseo.twitter.large

What is checked: The page uses the large card for X/Twitter.
Why it matters: Large cards are more visible and get more clicks.
Example: `<meta name="twitter:card" content="summary_large_image">`
Source: Open Graph protocol

Lowtwitter:image present (or og:image)seo.twitter.image

What is checked: An image is provided for X/Twitter, directly or as an og:image fallback.
Why it matters: Without an image there is no preview card.
Example: `<meta name="twitter:image" content="https://example.com/og.jpg">`
Source: Open Graph protocol

robots.txt

Mediumrobots.txt presentseo.robots.present

What is checked: Whether a file is served at `/robots.txt` with status 200.
Why it matters: Crawlers request robots.txt first. It controls which areas may be read and points to the sitemap.
Example: `User-agent: *`, `Allow: /`, `Sitemap: https://example.com/sitemap.xml`
Source: RFC 9309: Robots Exclusion Protocol

Lowrobots.txt syntax validseo.robots.syntax

What is checked: Whether robots.txt contains lines that are not a known directive.
Why it matters: Crawlers skip unknown lines, so a typo such as `Disalow` has no effect.
Example: `Disallow: /internal/` instead of `Disalow /internal/`
Source: Google Search Central: How Google interprets the robots.txt specification

Lowrobots.txt points to the sitemapseo.robots.sitemap

What is checked: Whether robots.txt contains a `Sitemap:` line.
Why it matters: This way every crawler finds the sitemap, including those you never submitted it to.
Example: `Sitemap: https://example.com/sitemap.xml`
Source: Google Search Central: Build and submit a sitemap

Lowrobots.txt has a wildcard ruleseo.robots.wildcard

What is checked: The robots.txt contains a `User-agent: *` group.
Why it matters: Without a general rule the default for all unlisted crawlers is missing.
Example: `User-agent: *` followed by `Allow: /`.
Source: RFC 9309 (robots.txt)

sitemap.xml

Mediumsitemap.xml presentseo.sitemap.present

What is checked: Whether a file is served at `/sitemap.xml` with status 200.
Why it matters: The sitemap lists every page that should be indexed. New and deeply linked pages are found faster.
Example: `<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9"><url><loc>https://example.com/</loc></url></urlset>`
Source: Google Search Central: Build and submit a sitemap

Mediumsitemap.xml is valid XMLseo.sitemap.valid

What is checked: Whether the sitemap starts with an XML declaration, `<urlset>` or `<sitemapindex>`.
Why it matters: Some servers return an HTML page at `/sitemap.xml` (such as the home page or an error page). Crawlers cannot use that.
Example: `<?xml version="1.0" encoding="UTF-8"?>` followed by `<urlset …>`
Source: sitemaps.org: Sitemaps XML format

LowSitemap contains the checked URLseo.sitemap.contains

What is checked: Whether the checked page, or at least the domain, appears in the sitemap.
Why it matters: Pages missing from the sitemap are crawled less often and later.
Example: `<url><loc>https://example.com/services/</loc></url>`
Source: Google Search Central: Build and submit a sitemap

LowSitemap contains lastmodseo.sitemap.lastmod

What is checked: Whether the sitemap contains `<lastmod>` entries.
Why it matters: Google uses lastmod to recrawl changed pages sooner, as long as the date is accurate.
Example: `<lastmod>2026-10-01</lastmod>` only when the content really changed
Source: Google Search Central: Build and submit a sitemap

LowSitemap and indexability do not contradictseo.sitemap.indexconsistent

What is checked: A noindex page is not listed in the sitemap, an indexable page is.
Why it matters: Contradictory signals confuse search engines and cost trust.
Example: Remove the noindex page from the sitemap or drop noindex.
Source: sitemaps.org

MediumSitemap contains entriesseo.sitemap.entries

What is checked: The sitemap lists at least one address or a child index.
Why it matters: An empty sitemap does not help search engines.
Example: `<urlset><url><loc>https://example.com/</loc></url></urlset>`
Source: sitemaps.org

LowNo legal pages in the sitemapseo.sitemap.nolegal

What is checked: Imprint, privacy policy, terms, cookie notices and terms of use are not listed in the sitemap.
Why it matters: Legal pages exist for visitors, not as search results.
Example: Remove `/imprint` and `/privacy` from the sitemap.
Source: sitemaps.org

404 page

MediumUnknown URL returns a real 404seo.404.real

What is checked: Whether a deliberately non-existent address answers with status 404.
Why it matters: If an error page returns status 200 (soft 404), Google treats it as real content, and empty or duplicate pages end up in the index.
Example: `/does-not-exist` → `HTTP/2 404` with a helpful error page
Source: Google Search Central: Soft 404 errors

Low404 page has contentseo.404.content

What is checked: The 404 page explains the situation in a few sentences.
Why it matters: A bare error message leaves visitors stranded.
Example: “This page no longer exists. Here is the home page.”
Source: Google Search Central

Low404 page links back homeseo.404.homelink

What is checked: The 404 page contains a link to the home page.
Why it matters: A clear way back reduces bounce rate.
Example: `<a href="/">Back to home</a>`
Source: Google Search Central

URL structure

LowURL without capitals, underscores or session parametersseo.404.url

What is checked: Whether the URL path contains capital letters or underscores and whether session parameters appear in the URL.
Why it matters: URLs are case-sensitive, which creates duplicate addresses. Google recommends hyphens over underscores, and session parameters create endless variants.
Example: `/tax-advice/` instead of `/Tax_Advice/?sid=123`
Source: Google Search Central: URL structure best practices

LowTrailing slash used consistentlyseo.url.trailingslash

What is checked: Internal links and the page itself use the trailing slash the same way.
Why it matters: With and without a slash creates two addresses for one page.
Example: Either `/services` everywhere or `/services/` everywhere, not mixed.
Source: Google Search Central

LowURL is readableseo.url.readable

What is checked: The path consists of readable words, not IDs or hashes.
Why it matters: Readable addresses get more clicks and are easier to understand.
Example: Good: `/blog/crypto-tax-austria`. Bad: `/p/8f14e45f`.
Source: Google Search Central

Robots meta tag

Lowrobots directive presentseo.robotsmeta.present

What is checked: The page has a robots meta or an X-Robots-Tag.
Why it matters: Without it defaults apply; one line documents the intent.
Example: `<meta name="robots" content="index, follow, max-snippet:-1, max-image-preview:large">`
Source: Google Search Central

MediumNo nofollow in the robots metaseo.robotsmeta.nofollow

What is checked: The page does not tell crawlers to ignore its links.
Why it matters: Page-level `nofollow` devalues every internal link and slows indexing.
Example: Wrong: `index, nofollow`. Right: `index, follow`.
Source: Google Search Central

Links and navigation

LowLinks have descriptive anchor textseo.links.anchortext

What is checked: Every link carries text that describes its target.
Why it matters: “Click here” tells search engines and screen readers nothing about the target.
Example: Wrong: “Read more”. Right: “View all services”.
Source: Google Search Central

LowNavigation element presentseo.nav.present

What is checked: The page has a `<nav>` element or `role="navigation"`.
Why it matters: Semantic navigation is clearer for screen readers and crawlers.
Example: `<nav aria-label="Main navigation">…</nav>`
Source: W3C: WAI-ARIA

MediumInternal links presentseo.links.internal

What is checked: The page links at least three of its own targets.
Why it matters: Internal links spread relevance and help crawlers explore the site.
Example: Link home, services and contact, not only the home page itself.
Source: Google Search Central

MediumLinks are real anchor elementsseo.links.static

What is checked: Navigation and content use `<a href>` instead of JavaScript-only clicks.
Why it matters: Crawlers and AI bots only follow real links in the HTML.
Example: Wrong: `<span onclick="go()">`. Right: `<a href="/prices">`.
Source: Google Search Central

AI visibility (38 checks)

Snippet control

Mediummax-image-preview:large allowedgeo.snippet.imagepreview

What is checked: The robots meta or X-Robots-Tag allows large image previews.
Why it matters: Without it Google only shows small preview images.
Example: `max-image-preview:large`
Source: Google: Snippet-Steuerung

MediumNo data-nosnippet on the main contentgeo.snippet.datanosnippet

What is checked: Sections with `data-nosnippet` do not cover the main content.
Why it matters: Text marked this way does not appear in search results and AI answers.
Example: Wrong: `<main data-nosnippet>`. Right: mark only small side notes.
Source: Google: Snippet-Steuerung

SevereNo nosnippetgeo.snippet.nosnippet

What is checked: Whether the robots meta tag contains `nosnippet`.
Why it matters: With `nosnippet`, Google may not show any text excerpt, including in AI Overviews. The page can then hardly serve as a source for answers.
Example: `<meta name="robots" content="index, follow">` without `nosnippet`
Source: Google Search Central: Robots meta tag specifications

Mediummax-snippet not too smallgeo.snippet.maxsnippet

What is checked: Whether `max-snippet` in the robots meta tag limits the text excerpt to fewer than 50 characters.
Why it matters: A very small value leaves hardly any text for search results and AI answers. `-1` allows excerpts of any length.
Example: `<meta name="robots" content="max-snippet:-1">`
Source: Google Search Central: Robots meta tag specifications

Answer-ready content

MediumContent reachable without JavaScriptgeo.content.noscript

What is checked: The page delivers enough text in the HTML or a noscript hint.
Why it matters: Bots without JavaScript only see the raw HTML; an empty container stays empty.
Example: `<noscript><p>Please enable JavaScript …</p></noscript>` plus server-rendered text.
Source: Google Search Central

SevereAt least 300 words of textgeo.content.words

What is checked: How many words of visible text the page's HTML contains.
Why it matters: AI models cite pages that answer a question themselves. With little text there is hardly anything to cite. 300 words is a guideline, not a Google requirement.
Example: A service page with a description, process, pricing and common questions
Source: Google Search Central: Creating helpful, reliable, people-first content

SevereContent not rendered by JavaScript onlygeo.content.notjs

What is checked: Whether the raw HTML (before JavaScript runs) contains text, and how much text there is compared to inline scripts.
Why it matters: Many AI crawlers do not run JavaScript. Anything that only appears in the browser stays invisible to them.
Example: Server-side rendering or static pages instead of an empty `<div id="root"></div>`
Source: Google Search Central: Understand JavaScript SEO basics

LowQuestion headings with a direct answergeo.content.qa

What is checked: Whether headings phrased as questions are followed by more content. Pages without question headings pass.
Why it matters: AI search likes to quote the answer right below a matching question. A question without an answer below it helps no one.
Example: `<h2>How much is a first consultation?</h2><p>The first consultation is free and takes 30 minutes.</p>`
Source: Google Search Central: Creating helpful, reliable, people-first content

LowLists or tables presentgeo.content.lists

What is checked: Whether the page contains lists (`<ul>`, `<ol>`) or tables (`<table>`).
Why it matters: Lists and tables are easy to read and to reuse in answers, such as steps, prices or comparisons.
Example: `<ol><li>Send documents</li><li>First meeting</li><li>Quote</li></ol>`
Source: MDN: <ul> element

LowConcrete facts in the textgeo.content.facts

What is checked: The text names numbers, years, percentages or names instead of general claims.
Why it matters: Concrete facts make content quotable for AI answers.
Example: “Since 2014 we have served 120 clients in Vienna.”
Source: Google Search Central

LowNo filler textgeo.content.filler

What is checked: The text avoids empty phrases and filler.
Why it matters: Filler dilutes the answer and is not quoted.
Example: Wrong: “In today's fast-paced world it is important to note …”.
Source: Google Search Central

LowAuthor identifiablegeo.content.author

What is checked: On article pages an author is named, in the schema, a meta tag or visibly.
Why it matters: AI systems prefer content with clear authorship.
Example: `<meta name="author" content="Max Muster">` or `author` in the article schema.
Source: Google Search Central

llms.txt

Mediumllms.txt presentgeo.llms.present

What is checked: Whether a file with content is served at `/llms.txt`.
Why it matters: llms.txt is a proposed short overview in Markdown: who you are and which pages matter. AI assistants find the right content faster.
Example: `# Example Ltd` + `> Tax advice in Graz` + `- [Services](https://example.com/services/)`
Source: llmstxt.org: The /llms.txt file

Lowllms.txt with H1 and short descriptiongeo.llms.structure

What is checked: Whether llms.txt starts with a `# heading` and contains a descriptive paragraph.
Why it matters: The format requires only the H1, followed by a short summary. Without it, a model lacks the context for the links.
Example: `# Example Ltd` followed by `> Tax advice for freelancers and small businesses in Graz.`
Source: llmstxt.org: The /llms.txt file

Lowllms.txt contains linksgeo.llms.links

What is checked: Whether llms.txt contains Markdown links `[name](URL)`.
Why it matters: The links show a model which pages hold the important answers. Without links the file is just a profile.
Example: `- [Pricing](https://example.com/pricing/): packages and hourly rates`
Source: llmstxt.org: The /llms.txt file

Lowllms-full.txt present (bonus)geo.llms.full

What is checked: Whether a file with the full content is served at `/llms-full.txt`. The file is optional: if it is missing, the check does not count.
Why it matters: Some tools load llms-full.txt to read all content in one piece instead of fetching every linked page separately.
Example: `/llms-full.txt` with the text of every page linked in llms.txt, as Markdown
Source: llmstxt.org: The /llms.txt file

Lowllms.txt linked in the headgeo.llms.linked

What is checked: The head contains an alternate link with `type="text/plain"` or `text/markdown"` to the llms.txt.
Why it matters: AI tools find the file without guessing.
Example: `<link rel="alternate" type="text/plain" href="/llms.txt">`
Source: llmstxt.org

Lowllms.txt starts with name and short descriptiongeo.llms.intro

What is checked: The file starts with an H1 and a quoted paragraph describing the company.
Why it matters: AI tools immediately recognise who runs the site.
Example: `# Muster GmbH` and below `> Muster GmbH advises small businesses in Vienna.`
Source: llmstxt.org

Lowllms.txt is plain textgeo.llms.plaintext

What is checked: llms.txt is served as `text/plain` or `text/markdown` and contains no HTML or JSON.
Why it matters: Only then can AI tools read the file directly.
Example: Response header `Content-Type: text/plain; charset=utf-8`.
Source: llmstxt.org

Lowllms.txt under 100 KBgeo.llms.size

What is checked: The file is smaller than 100 KB and was read completely.
Why it matters: Very large files get truncated by tools.
Example: Include only the most important pages and sections.
Source: llmstxt.org

AI crawlers in robots.txt

SevereAI search bots not blockedgeo.crawlers.search

What is checked: Whether robots.txt blocks one of the bots AI search tools use to read pages for answers: GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, Claude-SearchBot, PerplexityBot, Google-Extended.
Why it matters: A blocked bot may not read the page, so it cannot appear as a source in ChatGPT, Claude or Perplexity.
Example: No `User-agent: GPTBot` group with `Disallow: /` in robots.txt
Source: OpenAI: Overview of OpenAI crawlers

LowAI training botsgeo.crawlers.training

What is checked: Whether robots.txt blocks bots that collect content only for training models (Common Crawl's CCBot, Applebot-Extended).
Why it matters: Blocking is a deliberate choice and costs little: it keeps your texts out of training data, but AI search can still cite your page. Blocked training bots only cause a small deduction.
Example: `User-agent: CCBot` with `Disallow: /` blocks Common Crawl only.
Source: Common Crawl: CCBot

SevereOAI-SearchBot is allowedgeo.bots.oaisearchbot

What is checked: The rule checks whether robots.txt allows OAI-SearchBot to access the tested page.
Why it matters: OAI-SearchBot reads pages for search in ChatGPT.
Example: User-agent: OAI-SearchBot Allow: /
Source: OpenAI / Anthropic Crawler-Dokumentation

SevereChatGPT-User is allowedgeo.bots.chatgptuser

What is checked: The rule checks whether robots.txt allows ChatGPT-User to access the tested page.
Why it matters: ChatGPT-User fetches pages when someone mentions an address in chat.
Example: User-agent: ChatGPT-User Allow: /
Source: OpenAI / Anthropic Crawler-Dokumentation

SevereClaude-User is allowedgeo.bots.claudeuser

What is checked: The rule checks whether robots.txt allows Claude-User to access the tested page.
Why it matters: Claude-User fetches pages on behalf of a user.
Example: User-agent: Claude-User Allow: /
Source: OpenAI / Anthropic Crawler-Dokumentation

SevereClaude-SearchBot is allowedgeo.bots.claudesearchbot

What is checked: The rule checks whether robots.txt allows Claude-SearchBot to access the tested page.
Why it matters: Claude-SearchBot builds the search index for Claude.
Example: User-agent: Claude-SearchBot Allow: /
Source: OpenAI / Anthropic Crawler-Dokumentation

SeverePerplexityBot is allowedgeo.bots.perplexitybot

What is checked: The rule checks whether robots.txt allows PerplexityBot to access the tested page.
Why it matters: PerplexityBot supplies the sources for Perplexity answers.
Example: User-agent: PerplexityBot Allow: /
Source: OpenAI / Anthropic Crawler-Dokumentation

LowGPTBot is allowedgeo.bots.gptbot

What is checked: The rule checks whether robots.txt allows GPTBot to access the tested page.
Why it matters: GPTBot collects content for training OpenAI models.
Example: User-agent: GPTBot Allow: /
Source: OpenAI / Anthropic Crawler-Dokumentation

LowClaudeBot is allowedgeo.bots.claudebot

What is checked: The rule checks whether robots.txt allows ClaudeBot to access the tested page.
Why it matters: ClaudeBot collects content for training Anthropic models.
Example: User-agent: ClaudeBot Allow: /
Source: OpenAI / Anthropic Crawler-Dokumentation

LowGoogle-Extended is allowedgeo.bots.googleextended

What is checked: The rule checks whether robots.txt allows Google-Extended to access the tested page.
Why it matters: Google-Extended controls whether Google uses content for Gemini.
Example: User-agent: Google-Extended Allow: /
Source: OpenAI / Anthropic Crawler-Dokumentation

Entity

MediumOrganization/Person JSON-LD presentgeo.entity.identity

What is checked: Whether the page contains JSON-LD of type Organization, Person, ProfessionalService or LocalBusiness.
Why it matters: It tells machines unambiguously who is behind the site. Search engines and AI models then attribute content to the right business.
Example: `{"@context":"https://schema.org","@type":"Organization","name":"Example Ltd","url":"https://example.com"}`
Source: Google Search Central: Organization structured data

MediumsameAs with at least 2 profilesgeo.entity.sameas

What is checked: How many profiles (LinkedIn, X, GitHub, company register …) the entity schema links via `sameAs`.
Why it matters: sameAs connects your website with your other profiles. This helps to tell you apart from others with the same name.
Example: `"sameAs": ["https://www.linkedin.com/company/example", "https://github.com/example"]`
Source: schema.org: sameAs

LowEntity with name, url, logogeo.entity.logo

What is checked: Whether an Organization schema contains name, url and logo together.
Why it matters: With these three properties Google can show your logo in the knowledge panel and in search results.
Example: `"name": "Example Ltd", "url": "https://example.com", "logo": "https://example.com/logo.png"`
Source: Google Search Central: Organization structured data

LowknowsAbout in the entity schemageo.entity.knowsabout

What is checked: The Organization or Person schema names topics in `knowsAbout`.
Why it matters: This helps search engines and AI assistants place the entity by topic.
Example: `"knowsAbout": ["Taxes", "Accounting"]`
Source: Schema.org

LowServices marked up as Offergeo.entity.offers

What is checked: The entity schema describes services with `makesOffer` or `hasOfferCatalog`.
Why it matters: This tells Google which services are offered.
Example: `"makesOffer": [{"@type":"Offer","itemOffered":{"@type":"Service","name":"Consulting"}}]`
Source: Google: Strukturierte Daten

LowPerson links a LinkedIn profilegeo.entity.personlinkedin

What is checked: A person in the schema references a personal LinkedIn profile via sameAs.
Why it matters: The profile substantiates the identity of the person behind the page.
Example: `"sameAs": ["https://www.linkedin.com/in/max-muster"]`
Source: Schema.org

Freshness

LowFreshness signal presentgeo.freshness

What is checked: Whether the page states a modification date: the `Last-Modified` header, `article:modified_time` or `dateModified`.
Why it matters: For questions about prices, laws or deadlines, search engines and AI models prefer current sources. Without a date, freshness cannot be told.
Example: `<meta property="article:modified_time" content="2026-10-01T09:00:00+02:00">`
Source: Google Search Central: Influence the byline date

LowVisible date on article pagesgeo.content.dateshown

What is checked: The article page shows a publication or modification date in the text.
Why it matters: A visible date makes freshness verifiable.
Example: `<time datetime="2026-10-12">12 October 2026</time>`
Source: Google Search Central

Structured data (29 checks)

JSON-LD general

MediumAt least one JSON-LD blockstructured.jsonld.present

What is checked: The page contains at least one block of structured data as JSON-LD.
Why it matters: Structured data helps search engines understand content and entities.
Example: `<script type="application/ld+json">{"@context":"https://schema.org","@type":"Organization"}</script>`
Source: Google: Strukturierte Daten

SevereAll JSON-LD blocks are parseablestructured.jsonld.parseable

What is checked: Every JSON-LD block is valid JSON.
Why it matters: A syntax error makes every statement in the block unusable.
Example: Check the block with a JSON validator.
Source: Google: Strukturierte Daten

Medium@context schema.orgstructured.jsonld.context

What is checked: The blocks carry `@context: https://schema.org`.
Why it matters: Without a context the types are ambiguous for search engines.
Example: `"@context": "https://schema.org"`
Source: Schema.org

Low@id with an absolute URLstructured.entity.id

What is checked: Whether the main entities (Organization, Person, WebSite) carry an @id that is an absolute URL.
Why it matters: A stable @id makes an entity recognisable across pages. Other blocks can point to it instead of repeating the same details.
Example: `{"@type":"Organization","@id":"https://example.com/#organization","name":"Example Ltd"}`
Source: W3C JSON-LD 1.1: Node identifiers

LowinLanguage declaredstructured.inlanguage

What is checked: Whether the WebSite, WebPage or article states the content language with inLanguage, and whether it matches the page's lang attribute.
Why it matters: The language tag helps match the content to the right language version and audience, which matters most on multilingual sites.
Example: `{"@type":"WebSite","name":"Example Ltd","url":"https://example.com/","inLanguage":"en-GB"}`
Source: schema.org: inLanguage

LowSchema image URLs are absolutestructured.image.absolute

What is checked: Whether every image reference in the JSON-LD (image, logo, thumbnailUrl) is a full http(s) address.
Why it matters: Structured data is often processed apart from the page. Relative paths or data URIs cannot be resolved reliably there, and the image goes missing from the result.
Example: `"logo": "https://example.com/logo.png"` instead of `"logo": "/logo.png"`
Source: Google Search Central: General structured data guidelines

WebSite

LowWebSite schema presentstructured.website.present

What is checked: A WebSite schema describes the site as a whole.
Why it matters: It links pages to name, URL and language.
Example: `{"@type":"WebSite","name":"Example","url":"https://example.com"}`
Source: Google: Strukturierte Daten

LowWebSite with name and urlstructured.website.fields

What is checked: The WebSite schema contains `name` and `url`.
Why it matters: Both are required for the markup.
Example: `{"@type":"WebSite","name":"Example","url":"https://example.com"}`
Source: Google: Strukturierte Daten

Breadcrumbs

Articles

MediumArticle schema complete (for articles)structured.article

What is checked: Articles name `headline`, `author`, `datePublished` and `image`.
Why it matters: This is the basis for article rich results.
Example: `{"@type":"Article","headline":"…","author":{…},"datePublished":"2026-10-12","image":"https://…"}`
Source: Google: Strukturierte Daten

Other schema types

LowOther schema types completestructured.other

What is checked: FAQPage, Product, LocalBusiness, Event, Review and VideoObject are complete when present.
Why it matters: Incomplete blocks produce no rich results.
Example: Add the required fields per type from the Google documentation.
Source: Google: Strukturierte Daten

Organization and person

LowOrganization with name, url, address and sameAsstructured.org.fields

What is checked: Whether the Organization schema (including subtypes such as LocalBusiness or ProfessionalService) states a name, the website URL, a postal address and sameAs profiles.
Why it matters: These details let Google tie the page to your company without guessing, for example in the knowledge panel. sameAs links the site to your profiles elsewhere.
Example: `{"@type":"Organization","name":"Example Ltd","url":"https://example.com/","address":{"@type":"PostalAddress","streetAddress":"1 Main Street","postalCode":"1010","addressLocality":"Vienna","addressCountry":"AT"},"sameAs":["https://www.linkedin.com/company/example"]}`
Source: Google Search Central: Organization structured data

LowPerson schemastructured.person.present

What is checked: Whether the structured data describes a person, for example as an author, an owner or the subject of a profile page. Its absence does not count against you.
Why it matters: For consultants, authors and sole traders, a Person schema helps search engines and AI assistants attribute content to a specific person.
Example: `{"@type":"Person","name":"Anna Example","jobTitle":"Tax adviser","url":"https://example.com/team/anna"}`
Source: schema.org: Person

FAQPage

LowFAQPage schemastructured.faq.present

What is checked: Whether the page has a FAQPage schema. Its absence does not count; it only makes sense on pages with real questions and answers.
Why it matters: Since 2023 Google shows FAQ results only for well-known government and health sites. Clearly marked-up questions and answers still help search engines and AI assistants find and quote your answers.
Example: `{"@type":"FAQPage","mainEntity":[{"@type":"Question","name":"What does the first meeting cost?","acceptedAnswer":{"@type":"Answer","text":"The first meeting is free."}}]}`
Source: Google Search Central: FAQPage structured data

MediumEvery FAQ question has an answerstructured.faq.answers

What is checked: Whether every Question in the FAQPage schema has a name and an acceptedAnswer with text.
Why it matters: name and acceptedAnswer.text are required. A question without an answer is useless to search engines and makes the markup invalid.
Example: `"acceptedAnswer": {"@type":"Answer","text":"We reply within one working day."}`
Source: Google Search Central: FAQPage structured data

MediumFAQ content visible on the pagestructured.faq.visible

What is checked: Whether each question from the FAQPage schema, and roughly its answer, appears in the visible page text. Text in scripts or hidden with the hidden attribute or display:none does not count.
Why it matters: Google requires marked-up questions and answers to be visible to visitors. Markup for content nobody can see is treated as misleading and can lead to a manual action.
Example: Questions as headings or `<summary>`, with the answer as text right below.
Source: Google Search Central: FAQPage structured data

Ratings

LowAggregateRatingstructured.rating.present

What is checked: Whether the structured data contains an overall rating (AggregateRating), including one nested in Product, LocalBusiness or similar types. Its absence does not count.
Why it matters: Genuine ratings can appear as stars in search results. Google only allows this for certain types, and not for ratings a business publishes about itself on its own site.
Example: `"aggregateRating":{"@type":"AggregateRating","ratingValue":4.7,"reviewCount":38,"bestRating":5}`
Source: Google Search Central: Review snippet structured data

LowRating with ratingValue and a countstructured.rating.fields

What is checked: Whether every AggregateRating has a ratingValue and a count (reviewCount or ratingCount).
Why it matters: Both are required. Without a count the rating cannot be put in context, and Google shows no stars.
Example: `"ratingValue": 4.7, "reviewCount": 38`
Source: Google Search Central: Review snippet structured data

LowbestRating declaredstructured.rating.best

What is checked: Whether the aggregate rating states its scale with bestRating, and whether ratingValue falls within it.
Why it matters: Without bestRating Google assumes a scale up to 5. If you rate on another scale, such as out of 10, the value is shown wrongly.
Example: `"ratingValue": 9.1, "bestRating": 10, "worstRating": 1`
Source: Google Search Central: Review snippet structured data

Events

LowEvent schemastructured.event.present

What is checked: Whether the page marks up events as Event (or a subtype such as BusinessEvent). Its absence does not count.
Why it matters: Google can show marked-up events with date and place in its event search. That is worth it for seminars, webinars and other dates.
Example: `{"@type":"Event","name":"Tax update 2027","startDate":"2027-01-20T18:00+01:00","location":{"@type":"Place","name":"Example Ltd","address":"1 Main Street, 1010 Vienna"}}`
Source: Google Search Central: Event structured data

LowEvent with name and startDatestructured.event.fields

What is checked: Whether every event has a name and a start date in ISO 8601 format.
Why it matters: name and startDate are required. A date like "20 January" cannot be read reliably by a machine; with time and time zone the slot is unambiguous.
Example: `"startDate": "2027-01-20T18:00+01:00"`
Source: Google Search Central: Event structured data

LowEvent location with addressstructured.event.location

What is checked: Whether every event has a location: a Place with an address or, for online events, a VirtualLocation.
Why it matters: Google requires location. For in-person events the Place needs an address, otherwise the event cannot be matched to local searches.
Example: `"location":{"@type":"Place","name":"Seminar room","address":{"@type":"PostalAddress","streetAddress":"1 Main Street","addressLocality":"Vienna","postalCode":"1010","addressCountry":"AT"}}`
Source: Google Search Central: Event structured data

Video

LowVideoObjectstructured.video.present

What is checked: Whether videos on the page are marked up as VideoObject. Its absence does not count.
Why it matters: VideoObject tells Google what a video is about, so it can show it with a thumbnail in search and in the video tab.
Example: `{"@type":"VideoObject","name":"How the import works","description":"Two minutes on the CSV import.","thumbnailUrl":"https://example.com/video.jpg","uploadDate":"2026-10-01T09:00:00+02:00"}`
Source: Google Search Central: Video structured data

LowVideo has a namestructured.video.name

What is checked: Whether every VideoObject has a title (name).
Why it matters: name is required. Without a title the video is not shown as a video result.
Example: `"name": "How the import works"`
Source: Google Search Central: Video structured data

LowVideo has a descriptionstructured.video.description

What is checked: Whether every VideoObject has a description.
Why it matters: Google recommends a description. It tells search engines and AI assistants what the video shows without playing it.
Example: `"description": "Two minutes on importing a CSV from your accounts."`
Source: Google Search Central: Video structured data

LowVideo has a thumbnailUrlstructured.video.thumbnail

What is checked: Whether every VideoObject provides a preview image through thumbnailUrl.
Why it matters: thumbnailUrl is required. Without a thumbnail Google does not show the video as a video result.
Example: `"thumbnailUrl": ["https://example.com/video-16x9.jpg"]`
Source: Google Search Central: Video structured data

LowVideo has an uploadDatestructured.video.uploaddate

What is checked: Whether every VideoObject has a publication date in ISO 8601 format, ideally with time and time zone.
Why it matters: uploadDate is required. Google recommends adding the time zone; otherwise the time is read in Googlebot's time zone.
Example: `"uploadDate": "2026-10-01T09:00:00+02:00"`
Source: Google Search Central: Video structured data

Technology (62 checks)

Mobile

SevereViewport meta presenttech.viewport

What is checked: The page sets `<meta name="viewport">`.
Why it matters: Without a viewport mobile browsers zoom out and the page looks tiny.
Example: `<meta name="viewport" content="width=device-width, initial-scale=1">`
Source: Google: Mobile-freundliche Seiten

Security headers

MediumContent-Security-Policy settech.security.csp

What is checked: The `Content-Security-Policy` header is set.
Why it matters: A CSP limits which scripts and resources may load.
Example: `Content-Security-Policy: default-src 'self'`
Source: MDN: HTTP headers

LowX-Content-Type-Options settech.security.xcto

What is checked: The `X-Content-Type-Options: nosniff` header is set.
Why it matters: It stops browsers from interpreting files as another type.
Example: `X-Content-Type-Options: nosniff`
Source: MDN: HTTP headers

LowReferrer-Policy settech.security.referrer

What is checked: The `Referrer-Policy` header is set.
Why it matters: It limits which addresses are passed to other sites.
Example: `Referrer-Policy: strict-origin-when-cross-origin`
Source: MDN: HTTP headers

LowPermissions-Policy settech.security.permissions

What is checked: The `Permissions-Policy` header is set.
Why it matters: It disables browser features such as camera and microphone.
Example: `Permissions-Policy: camera=(), microphone=(), geolocation=()`
Source: MDN: HTTP headers

LowClickjacking protectiontech.security.clickjacking

What is checked: The page protects itself with `frame-ancestors` or `X-Frame-Options`.
Why it matters: Otherwise the page can be embedded invisibly in another site.
Example: `frame-ancestors 'none'` in the CSP or `X-Frame-Options: DENY`.
Source: MDN: HTTP headers

LowCross-Origin-Opener-Policy settech.security.coop

What is checked: The `Cross-Origin-Opener-Policy` header is set.
Why it matters: It isolates browsing contexts and makes window-based attacks harder.
Example: `Cross-Origin-Opener-Policy: same-origin`
Source: MDN: HTTP headers

LowCross-Origin-Embedder-Policy settech.security.coep

What is checked: The `Cross-Origin-Embedder-Policy` header is set.
Why it matters: It prevents embedding foreign resources without permission.
Example: `Cross-Origin-Embedder-Policy: credentialless` (or `require-corp` if no third-party resources are needed)
Source: MDN: HTTP headers

LowCross-Origin-Resource-Policy settech.security.corp

What is checked: The `Cross-Origin-Resource-Policy` header is set.
Why it matters: It limits who may embed your files.
Example: `Cross-Origin-Resource-Policy: same-site`
Source: MDN: HTTP headers

Caching and compression

MediumBrotli or gzip enabledtech.caching.compression

What is checked: Text responses are compressed with Brotli, gzip or zstd.
Why it matters: Compression saves transfer time and bandwidth.
Example: The response header states `content-encoding: br` or `gzip`.
Source: MDN: HTTP caching

LowCache-Control for static filestech.caching.cachecontrol

What is checked: Static files carry a long-lived `Cache-Control`.
Why it matters: Repeat visits then load the files from cache.
Example: `Cache-Control: public, max-age=31536000, immutable`
Source: MDN: HTTP caching

LowHTML size under 500 KBtech.caching.htmlsize

What is checked: The HTML response stays under 500 KB.
Why it matters: Large HTML documents delay rendering and parsing.
Example: Move inline CSS and JS into separate files.
Source: web.dev

MediumCache-Control on the main responsetech.caching.main

What is checked: The HTML response carries a `Cache-Control` header.
Why it matters: Without it, browsers and CDNs decide inconsistently.
Example: `Cache-Control: public, max-age=0, must-revalidate`
Source: MDN: HTTP caching

LowPublic page is cacheabletech.caching.public

What is checked: The response does not forbid caching unnecessarily.
Why it matters: `no-store` or `private` slow down CDN and browser.
Example: `Cache-Control: public, max-age=300` instead of `no-store`
Source: MDN: HTTP caching

LowETag presenttech.caching.etag

What is checked: The response carries an `ETag`.
Why it matters: Crawlers then avoid re-downloading unchanged pages.
Example: `ETag: "5f3c9a"`
Source: MDN: HTTP caching

Images

Mediumalt on all content imagestech.images.alt

What is checked: All content images have an `alt` attribute.
Why it matters: Screen readers and image search need the alternative text.
Example: `<img src="team.jpg" alt="Team in a consultation">`
Source: Google: Bilder

LowModern image formats (WebP/AVIF)tech.images.format

What is checked: Raster images are served as WebP or AVIF; SVG is exempt.
Why it matters: Modern formats are much smaller at the same quality.
Example: `team.webp` instead of `team.jpg`.
Source: Google: Bilder

Lowwidth and height settech.images.dimensions

What is checked: Images declare `width` and `height`.
Why it matters: This reserves space and prevents layout shifts.
Example: `<img src="team.webp" width="800" height="600" alt="…">`
Source: web.dev

Lowloading=lazy below the foldtech.images.lazy

What is checked: Below-the-fold images use `loading="lazy"`.
Why it matters: This saves bandwidth on the first load.
Example: `<img src="…" loading="lazy" alt="…">` from the second image onwards.
Source: web.dev

LowHero image is not lazy-loadedtech.images.herolazy

What is checked: The most important visible image loads immediately; an image with fetchpriority=high must not be lazy either.
Why it matters: A lazy-loaded hero image delays the visible paint (LCP).
Example: `<img src="hero.jpg" fetchpriority="high">` without `loading="lazy"`.
Source: web.dev

LowHero image uses fetchpriority=hightech.images.fetchpriority

What is checked: The first content image carries `fetchpriority="high"`, directly or via a preload in the head.
Why it matters: The browser then loads it before other resources.
Example: `<img src="hero.jpg" fetchpriority="high">` or `<link rel="preload" as="image" href="hero.jpg" fetchpriority="high">`.
Source: web.dev

MediumAlt texts describe the imagetech.images.altquality

What is checked: Alt texts describe the image and repeat neither file names nor placeholders.
Why it matters: Descriptive alt texts help screen readers and image search.
Example: Wrong: `alt="IMG_1234.jpg"`. Right: `alt="Team in a consultation"`.
Source: Google: Bilder

LowImages under 200 KBtech.images.weight

What is checked: The sampled images are smaller than about 200 KB.
Why it matters: Large images slow the page, especially on mobile.
Example: Compress images or serve smaller variants via srcset.
Source: web.dev

Load time

MediumTime to First Byte under 800 mstech.load.ttfb

What is checked: The server responds in less than 800 ms.
Why it matters: A fast first byte speeds up every following step.
Example: Caching, a CDN or less work in the backend.
Source: web.dev

LowBlocking scripts in the headtech.load.blockingscripts

What is checked: The head contains no blocking scripts without `async` or `defer`.
Why it matters: Blocking scripts hold up the first paint.
Example: `<script src="app.js" defer></script>`
Source: web.dev

LowCSS included in the headtech.load.cssinhead

What is checked: Stylesheets appear as `<link rel="stylesheet">` or `<style>` in the head.
Why it matters: CSS in the body delays the first paint and can shift content.
Example: `<link rel="stylesheet" href="/style.css">` in the `<head>`.
Source: web.dev

MediumResponse time under 3 secondstech.load.responsetime

What is checked: The server responds in less than three seconds.
Why it matters: Slow responses cost rankings and visitors.
Example: Cache the home page at the network edge (CDN).
Source: web.dev

MX and SPF

MediumSPF present and strict enoughtech.email.spf

What is checked: The domain has an SPF record ending in `-all` or `~all`.
Why it matters: SPF prevents foreign servers from sending mail in your name.
Example: `v=spf1 include:… -all`
Source: Google Search Central

LowMX records presenttech.email.mx

What is checked: The domain has at least one MX record.
Why it matters: Without MX the domain cannot receive email.
Example: `example.com MX 10 mx1.mailprovider.com.`
Source: Google Search Central

DMARC

MediumDMARC with p=quarantine or rejecttech.email.dmarc

What is checked: The DMARC record is set to `p=quarantine` or `p=reject`.
Why it matters: This protects the domain's reputation against spam and spoofing.
Example: `v=DMARC1; p=quarantine; rua=mailto:dmarc@example.com`
Source: Google Search Central

LowDMARC with a reporting addresstech.email.dmarc_rua

What is checked: The DMARC record states `rua=` for reports.
Why it matters: Without reports the policy's effect stays invisible.
Example: `rua=mailto:dmarc@example.com`
Source: Google Search Central

DKIM

LowDKIM foundtech.email.dkim

What is checked: A DKIM record is reachable on a common selector.
Why it matters: DKIM signs outgoing mail and proves its origin.
Example: `<selector>._domainkey` as TXT with `v=DKIM1; p=…`.
Source: Google Search Central

LowDKIM with a public keytech.email.dkim_key

What is checked: The DKIM record contains a key (`p=`).
Why it matters: Without a key no signature can be verified.
Example: `v=DKIM1; k=rsa; p=MIIBIjANBg…`
Source: Google Search Central

MTA-STS

LowMTA-STS presenttech.email.mtasts

What is checked: An MTA-STS policy is published on the host `mta-sts.<domain>`.
Why it matters: It requires encrypted delivery between mail servers.
Example: `https://mta-sts.example.com/.well-known/mta-sts.txt` with `version: STSv1`.
Source: RFC 8461 (MTA-STS)

LowMTA-STS in enforce modetech.email.mtasts_mode

What is checked: The MTA-STS policy is set to `mode: enforce`.
Why it matters: Only then do mail servers enforce encrypted delivery.
Example: `mode: enforce`
Source: RFC 8461 (MTA-STS)

Low_mta-sts is announcedtech.email.mtasts_dns

What is checked: A `_mta-sts` TXT record announces the policy.
Why it matters: Without it mail servers do not find the policy.
Example: `_mta-sts.example.com TXT "v=STSv1; id=20261018"`
Source: RFC 8461 (MTA-STS)

LowMTA-STS with max_agetech.email.mtasts_maxage

What is checked: The policy states `max_age`.
Why it matters: It determines how long mail servers cache the policy.
Example: `max_age: 604800`
Source: RFC 8461 (MTA-STS)

security.txt

Lowsecurity.txt with Contact and Expirestech.trust.securitytxt

What is checked: `/.well-known/security.txt` contains `Contact` and `Expires`.
Why it matters: Security researchers know where to send reports.
Example: `Contact: mailto:security@example.com` and `Expires: 2027-12-31T23:59:59.000Z`.
Source: RFC 9116 (security.txt)

Lowsecurity.txt with a valid Expirestech.trust.securitytxt_expires

What is checked: `Expires:` lies in the future.
Why it matters: An expired Expires makes the file invalid per RFC.
Example: `Expires: 2027-12-31T23:59:59.000Z`
Source: RFC 9116 (security.txt)

Lowsecurity.txt with Canonicaltech.trust.securitytxt_canonical

What is checked: `Canonical:` names the file location.
Why it matters: Multiple copies remain unambiguous.
Example: `Canonical: https://example.com/.well-known/security.txt`
Source: RFC 9116 (security.txt)

Lowsecurity.txt with Preferred-Languagestech.trust.securitytxt_languages

What is checked: `Preferred-Languages:` names the security team's languages.
Why it matters: Reporters know which language they can use.
Example: `Preferred-Languages: en, de`
Source: RFC 9116 (security.txt)

Trust

MediumImprint and privacy linktech.trust.legal

What is checked: The page links the imprint and the privacy policy.
Why it matters: This is mandatory in Austria and Germany; visitors look for it in the footer.
Example: Links “Imprint” and “Privacy” in the footer.
Source: Google Search Central

Analytics and consent

LowAnalytics detectedtech.analytics.detected

What is checked: The page includes Google Analytics, Tag Manager or a privacy-friendly tool.
Why it matters: Without measurement, decision data is missing; this is a hint, not an error.
Example: `gtag.js` with a measurement ID or a Tag Manager container.
Source: Google Search Central

MediumConsent is set before analyticstech.analytics.consentfirst

What is checked: The default consent appears in the HTML before the analytics loader.
Why it matters: Otherwise data is collected without consent (GDPR risk).
Example: `gtag('consent','default',{ad_storage:'denied',…})` before gtag.js loads.
Source: Google Search Central

LowAnalytics does not block renderingtech.analytics.async

What is checked: Analytics scripts use `async` or `defer`, or the Tag Manager loads them later.
Why it matters: A blocking script delays the first paint.
Example: `<script async src="…/gtag/js">`
Source: web.dev

MediumAll four consent types start as deniedtech.analytics.denied

What is checked: `ad_storage`, `analytics_storage`, `ad_user_data` and `ad_personalization` start as `denied`.
Why it matters: This is the privacy-friendly default; `granted` before consent is a risk.
Example: `gtag('consent','default',{ad_storage:'denied',analytics_storage:'denied',ad_user_data:'denied',ad_personalization:'denied'});`
Source: Google Search Central

Favicon and icons

LowFavicon presenttech.icons.favicon

What is checked: The page links an icon or serves `/favicon.ico`.
Why it matters: The icon appears in tabs and bookmarks.
Example: `<link rel="icon" href="/favicon.ico">`
Source: Google Search Central

LowSVG favicon linkedtech.icons.svg

What is checked: An icon with `type="image/svg+xml"` is linked.
Why it matters: SVG stays sharp at any size and is small.
Example: `<link rel="icon" href="/icon.svg" type="image/svg+xml">`
Source: Google Search Central

Lowapple-touch-icon linkedtech.icons.apple

What is checked: An apple-touch-icon is linked.
Why it matters: iOS uses it when adding to the home screen.
Example: `<link rel="apple-touch-icon" href="/apple-touch-icon.png">`
Source: Google Search Central

Web app manifest

LowWeb manifest linkedtech.manifest.present

What is checked: The page links a web app manifest.
Why it matters: The manifest describes name, icons and app start behaviour.
Example: `<link rel="manifest" href="/site.webmanifest">`
Source: MDN: Web app manifest

LowManifest is reachable and validtech.manifest.reachable

What is checked: The linked manifest responds with 200 and valid JSON.
Why it matters: A broken link causes errors and no app metadata.
Example: `{"name":"Example","icons":[…]}` as the manifest response.
Source: MDN: Web app manifest

LowManifest with a maskable icontech.manifest.maskable

What is checked: At least one icon in the manifest has `purpose` containing `maskable`.
Why it matters: Android otherwise crops icons awkwardly.
Example: `{"src":"/icon-512.png","purpose":"any maskable"}`
Source: MDN: Web app manifest

Lowtheme-color settech.themecolor

What is checked: The page sets `theme-color`.
Why it matters: Mobile browsers tint the address bar to match the brand.
Example: `<meta name="theme-color" content="#123456">`
Source: MDN: Web app manifest

Accessibility

DNS (CAA)

LowCAA record settech.email.caa

What is checked: The domain authorises certificate authorities via CAA.
Why it matters: CAA prevents foreign authorities from issuing certificates.
Example: `example.com CAA 0 issue "letsencrypt.org"`
Source: Google Search Central

Bot access

SeverePage reachable without a bot challengetech.bots.no_challenge

What is checked: Whether our request gets the page content or a firewall block page (Cloudflare, Sucuri, Akamai, DataDome, PerimeterX, Imperva).
Why it matters: Search engines and AI assistants fetch pages without a browser. If the firewall shows them a block page, the content is neither read nor cited.
Example: Configure the bot protection to let known crawlers (verified bots) through, for example allow "Verified Bots" in Cloudflare.
Source: Google Search Central: Overview of Google crawlers

SevereNo JavaScript challenge or CAPTCHA before the contenttech.bots.no_captcha

What is checked: Whether a CAPTCHA (reCAPTCHA, hCaptcha, Turnstile) or a check that needs JavaScript stands in front of the content.
Why it matters: Crawlers do not solve CAPTCHAs and often do not run JavaScript. A CAPTCHA in front of the whole page locks them out; inside a form it is fine.
Example: Use the CAPTCHA only in the form, not as an entry page.
Source: Google Search Central: JavaScript SEO basics

SevereGooglebot gets the contenttech.bots.googlebot

What is checked: We also fetch the page with the Googlebot token in the user agent (marked as a RELACON probe) and compare the result with our normal request.
Why it matters: Some firewalls block anything that calls itself Googlebot but does not come from Google's addresses; others let only Googlebot through. The comparison shows whether crawlers are treated differently.
Example: Use URL inspection in Search Console to test whether the real Googlebot can fetch the page.
Source: Google: Verifying Googlebot

MediumAI crawlers get the contenttech.bots.ai

What is checked: We fetch the page with the GPTBot token in the user agent (marked as a RELACON probe) and compare the result with our normal request.
Why it matters: Many bot protection services block AI crawlers by default, even when robots.txt allows them. Then the permission in robots.txt does not help.
Example: Allow AI crawlers in the firewall too, in Cloudflare under "AI Crawl Control".
Source: OpenAI: Overview of OpenAI crawlers

Mediumrobots.txt readable despite the firewalltech.bots.robots

What is checked: Whether /robots.txt is served as text and not as a firewall block page.
Why it matters: If a crawler gets a block page or a server error instead of robots.txt, Google treats the whole site as blocked to be safe.
Example: Exempt `/robots.txt` from the bot protection.
Source: Google: How Google interprets robots.txt

FAQ

Does the check store my page?

No content is stored beyond the report itself. The report contains the URL, the results and the time of the check.

Why does a rule show n/a?

n/a means the rule does not apply to this page, for example video rules on a page without video. n/a never costs points.

More about the website check

Check now