Skip to content
Metevix

The website intelligence platform

Your website has problemsyou can't see.

Metevix crawls every page, models how your URLs, links, content and crawlers actually relate, and ranks what to fix by the impact it will have. Technical SEO is one layer of it — performance, indexability, security and AI crawler access are the rest.

Analyze my websiteSee how it works

Free first audit. No credit card, no crawler install, no code changes.

checks per crawl
312checks per crawl
analysis domains
6analysis domains
crawl modes
6crawl modes
https://acme-commerce.com
Analyzing
  1. 01DiscoverSitemaps, feeds, internal links
  2. 02CrawlFetch + render, JS execution
  3. 03UnderstandDOM, schema, link graph, content
  4. 04Diagnose312 technical checks
  5. 05PrioritizeImpact × effort model
  6. 06RecommendExplained, actionable fixes

Website health

42 issues found across 1,284 pages

8 critical21 high13 medium

How it works

A website is data. Metevix makes it intelligence.

01 — WEBSITE

Website

Metevix starts from your site as it really is — every template, every route, every page that a crawler can reach.

8 templates · 1,284 routes

02 — CRAWLER

Crawler

The crawler walks the link graph, fetching and rendering each page with JavaScript executed, recording every response along the way.

1,284 fetched · 0 blocked

03 — DATA

Data

Markup, headers, timings, structured data and the full internal link graph are extracted into one connected model.

18,940 signals extracted

04 — ANALYSIS

Analysis

312 checks run against that model. Findings are bound to the exact URL and the exact evidence that produced them.

42 issues · 8 critical

05 — INTELLIGENCE

Intelligence

The result is not a report to read. It is a ranked, explained set of changes, scored by what each one will actually move.

Health 87 · 42 fixes ready

//services/products/blog/services/seo/services/audit/products/[id]/blog/[slug]87WEBSITE HEALTHTechnical SEO92Performance81Indexability95Security89

Website: Metevix starts from your site as it really is — every template, every route, every page that a crawler can reach. Crawler: The crawler walks the link graph, fetching and rendering each page with JavaScript executed, recording every response along the way. Data: Markup, headers, timings, structured data and the full internal link graph are extracted into one connected model. Analysis: 312 checks run against that model. Findings are bound to the exact URL and the exact evidence that produced them. Intelligence: The result is not a report to read. It is a ranked, explained set of changes, scored by what each one will actually move.

The crawl

It finds the pages you forgot you had.

Metevix follows every link, sitemap entry and redirect until the map is closed — then reports what came back, not what you expected.

Discoverycrawling…
  1. /200
  2. ├── /services200
  3. │ ├── /services/seo200
  4. │ └── /services/audit301
  5. ├── /products200
  6. │ └── /products/[id]200
  7. ├── /blog200
  8. │ └── /blog/[slug]200
  9. ├── /about200
  10. ├── /contact404
  11. └── /resources200

What the crawl resolved to

  • URLs discovered0
  • Indexable0
  • Pages with warnings0
  • Pages blocked or broken0
  • Quick wins available0

111 URLs were excluded from indexing — some deliberately, some not. Metevix separates the two, which is usually where the first real finding is.

The product

An audit that reads like an engineering report.

Metevix turns a full-site crawl into a single, prioritized view of what is wrong, how badly it matters, and what to do about it.

acme-commerce.comCrawl #48

Website health score

0issues

across 1,284 crawled pages

+6 since last crawl

Severity distribution

  • 8Critical
  • 21High
  • 13Medium

Analysis domains

  • Technical SEO92/100

    Canonicals, titles, structured data9 issues

  • Performance81/100

    Core Web Vitals across templates14 issues

  • Indexability95/100

    Robots, noindex, crawl budget3 issues

  • Security89/100

    HTTPS, headers, mixed content5 issues

  • Content78/100

    Duplication, thin pages, headings8 issues

  • Internal linking83/100

    Depth, orphans, anchor quality3 issues

One score, traceable to every finding behind it.

Each ring is an analysis domain. The score is weighted by how much each domain affects pages that actually earn traffic.

87/ 100
Technical SEO
92
Performance
81
Indexability
95
Security
89
Content
78
Internal linking
83

Website intelligence graph

Your site is a system, not a list of pages.

Metevix keeps one connected model of every URL — how it is linked, how deep it sits, what can reach it and what is wrong with it. Most findings only make sense in that context.

//products/blog/servicesorphaned

Hover or select a cluster

Depth, inbound links, crawlability and findings for any URL — and the cluster it belongs to. Orphaned pages are visible as the group nothing links into.

The same model answers questions a page-by-page report cannot: which templates concentrate link equity, which pages are reachable only from the sitemap, and which findings share a single upstream cause.

Technical SEO intelligence

Every finding comes with evidence.

Metevix does not hand you a list of warnings. Each issue names the affected URLs, explains the mechanism behind the problem, and gives you the exact change to make.

Critical37 pages

Missing canonical tags

What Metevix found
Product and filter URLs render without a canonical link element.
Why it matters
Google picks its own canonical for these URLs. Ranking signals split across parameter variants, and the wrong version can end up indexed.
What it is costing
22% of organic sessions land on URLs whose ranking signals are currently split across up to nine variants each.
Affected URLsshowing 4 of 37
  • /products/merino-crew?color=navy
  • /products/merino-crew?color=navy&size=m
  • /products/oxford-shirt?color=white
  • /products/oxford-shirt?fit=slim&color=white
How to fix it
Emit a self-referencing canonical on every indexable template, and point filtered variants at the clean parent URL.
Developer implementation
canonical.fixTypeScript
// app/products/[slug]/page.tsxexport async function generateMetadata({ params }) {  const { slug } = await params;  return {    alternates: {      canonical: `${SITE_URL}/products/${slug}`,    },  };}

Also detected

  • Broken links & 404s
  • Redirect chains & loops
  • Canonical conflicts
  • Unintended noindex
  • Robots.txt blocking
  • Sitemap errors
  • Hreflang clusters
  • Duplicate content
  • Orphan pages
  • Internal link depth
  • Structured data validity
  • Crawl budget waste
  • Pagination & faceting
  • Mixed content
  • Title & meta quality
  • Heading hierarchy

Crawler & AI readiness

Search engines are no longer the only thing reading your site.

Answer engines and model crawlers fetch pages under different rules, and most of them do not run JavaScript. Metevix reports what each one can actually reach.

Crawler access to this site, by agent: robots.txt rule, JavaScript rendering, and share of content reachable.
AgentPurposerobots.txtRenders JSContent reachable
GooglebotGoogleGoogle Search indexAllowedYes94%Renders JavaScript. 6% of URLs unreachable behind a noindex directive.
BingbotMicrosoftBing index, and Copilot groundingAllowedYes94%Same coverage as Googlebot on this site.
GPTBotOpenAITraining data collectionBlockedNo0%Disallowed in robots.txt. Intentional on this site — flagged so it stays a decision, not an accident.
OAI-SearchBotOpenAISurfacing pages in ChatGPT search resultsAllowedNo61%Does not execute JavaScript. 39% of content only exists after hydration.
ClaudeBotAnthropicTraining data collectionPartialNo44%Allowed on marketing paths, disallowed on /account and /checkout.
PerplexityBotPerplexityAnswer citationsAllowedNo61%Server-rendered content only, same 39% gap as other non-rendering agents.

Metevix reports what each crawler can reach on your site, based on your robots.txt, response codes and rendering. It does not claim to influence how any engine or model ranks, cites or weights that content.

Performance intelligence

Core Web Vitals, traced to their cause.

Field and lab measurements are correlated with what Metevix found in the page itself, so a failing metric always points at the resource responsible for it.

LCPLargest Contentful Paint

−56%
Before4.1s
After1.8s

Hero image served unoptimised and discovered late.

INPInteraction to Next Paint

−67%
Before420ms
After140ms

Long tasks from a blocking third-party bundle.

CLSCumulative Layout Shift

−92%
Before0.24
After0.02

Unreserved space for late-loading embeds.

TTFBTime to First Byte

−75%
Before1.24s
After0.31s

Uncached origin responses on template routes.

Illustrative figures from a representative e-commerce audit. Your results depend on your stack and templates.

  • JavaScript impact

    Long tasks are attributed to the bundles and third-party tags that caused them, with the parse, compile and execute cost of each broken out.

  • Image delivery

    Every image is checked for format, dimensions, compression and loading strategy — including whether the LCP candidate is discoverable in the initial HTML.

  • Resource loading

    Render-blocking requests, missing preloads, unused CSS and font-loading strategy are analysed across templates, not just the homepage.

  • Per-template diagnosis

    Results are grouped by template rather than by URL, so one fix in a layout resolves the issue on thousands of pages at once.

The intelligence layer

Data is cheap. Knowing what to do with it is not.

  1. 01

    Raw data

    1.2M DOM nodes

    Every response, header, byte and rendered node the crawl saw — stored exactly as it was served, not summarised.

  2. 02

    Signals

    18,940 extracted

    Canonicals, status codes, timings, schema, anchor text and link depth, normalised into facts that can be compared to each other.

  3. 03

    Patterns

    312 checks

    Those facts are evaluated across templates and against how search engines actually behave — not against a generic checklist.

  4. 04

    Problems

    42 findings

    Only the patterns that measurably cost you crawling, indexing or ranking survive. The rest stays out of your way.

  5. 05

    Priorities

    8 critical first

    Findings are ordered by pages affected, traffic at stake and the effort each fix will realistically take.

  6. 06

    Actions

    42 changes ready

    Each one written against your framework and your markup, ready to review, test and ship.

AI recommendations

Watch a problem become a fix.

Every finding carries its own resolution. The issue, the reason it matters, and the change to make are one continuous object — not three disconnected screens.

Detected
Critical

Missing canonical

37 pages affected · /products/[id]

Impact intelligence

Not what is broken. What to do on Monday.

Every finding is scored on the URLs it touches, the traffic on those URLs and the work it takes to resolve. The list you get back is already a plan.

Detected

0issues

A number alone is noise. Metevix resolves every finding into a tier with an expected outcome, so the list is already a plan.

  1. CriticalDo this week

    8

    Blocking indexing or breaking revenue pages. Every hour these stay live costs traffic.

  2. High impactDo this sprint

    21

    Measurably suppressing performance or consolidating signals onto the wrong URLs.

  3. MediumSchedule

    13

    Worth fixing, but no page is currently losing rankings because of them.

Fix first — ranked by modelled impactseverity × URLs × traffic ÷ effort

  1. Six pages cannot be indexed at all, and they sit on the template that earned 31% of organic sessions last quarter. One directive removal recovers all of it.

Ranked by modelled impact, not by severity. A critical issue on six unvisited pages sits below a high-severity one on the template that carries your traffic.

The loop

Most tools stop at detection.

Finding an issue is the cheap part. Metevix carries each one through to a verified fix — and re-opens it if the change did not do what it was supposed to.

  1. 01

    Detect

    A crawl finds the issue and binds it to the exact URL, response and markup that produced it.

    42 findings

  2. 02

    Explain

    Each finding is written out in plain language: what it is, and why it matters on your site specifically.

    Plain-language write-up

  3. 03

    Recommend

    A concrete change is proposed against your framework and your markup — not a generic best practice.

    Implementation diff

  4. 04

    Fix

    You apply the change. Metevix records what was expected to move, and by how much.

    Expected delta recorded

  5. 05

    Re-crawl

    The affected URLs are crawled again — just those, not the whole site — as soon as you mark the fix as shipped.

    Targeted re-crawl

  6. 06

    Verify

    The finding is confirmed resolved, or reopened with the reason it did not take. Nothing closes on assumption.

    Confirmed or reopened

Verified findings close themselves. Anything that fails verification returns to the top of the list with the reason it did not take — so the backlog reflects reality rather than what someone remembered to tick off.

Crawl modes

Run the analysis the question needs.

A pre-deploy check and a quarterly deep audit are not the same job. Metevix separates them, so you are never waiting on a full render pass to find out whether a redirect broke.

Fast crawl

under 2 min

Status codes, redirects, canonicals and indexability across every URL. The check you run before a deploy.

Runs

  • Response codes
  • Redirect chains
  • Canonicals
  • Indexability
  • Sitemap parity

Skips

  • JavaScript rendering
  • Performance timings
  • Content analysis

Page-level intelligence

Or skip the crawl entirely and analyze one URL.

Paste a single address and get the same checks against that page alone — markup, rendering, performance, structured data and crawler access — in a few seconds, with no project and no site-wide crawl.

Analyze a single page

Continuous monitoring

An audit is a snapshot. Sites change on Tuesday.

Metevix re-crawls on a schedule and after every deploy, compares each run with the last, and tells you the moment something regresses — rather than at the next quarterly audit.

Change feed — acme-commerce.comWatching
  1. New 5xx responses12 min ago34 URLs on /products began returning 503. First seen 11 minutes after the 14:20 deploy.Regression
  2. robots.txt changed2 hours agoA new Disallow: /blog rule appeared. 214 previously crawlable URLs are now blocked.Regression
  3. Canonical fix verifiedYesterdayAll 37 product variants now emit a self-referencing canonical. Finding closed automatically.Resolved
  4. LCP degraded on /products2 days agoMedian LCP moved 1.8s → 3.1s. A new 240KB script became render-blocking.Regression
  5. Indexable URL count changed4 days ago1,173 → 1,208 indexable URLs after a sitemap update. No issues introduced.Change

Alerts fire on new 5xx responses, indexability and robots changes, performance regressions, redirect breakage and lost structured data.

Three sources, one model

A crawl tells you what is on the site. Logs tell you what crawlers did with it. Search Console tells you whether any of it landed. Isolated, each is a report; joined on the URL, they answer questions none of them can alone.

  1. CrawlAvailable now

    What your site actually serves: markup, headers, timings, redirects and the full link graph.

    What is on the site, and what state is it in?

  2. Server logsOn the roadmap

    What crawlers genuinely requested, how often, and what your origin returned to them.

    Where is crawl budget actually going?

  3. Search ConsoleOn the roadmap

    What was indexed, impressed and clicked — joined to the URLs the crawl already knows about.

    Did any of it reach anyone?

Who it is for

The same crawl. Three different jobs.

Metevix produces one authoritative model of your site. What changes is the view onto it, and the language it is explained in.

Developers

Catch it before it ships

Run Metevix against a preview deployment and see exactly which template introduced a regression, with the failing markup quoted next to the rule that caught it.

  • Crawl preview and staging environments
  • Findings grouped by template, not by URL
  • Framework-aware implementation guidance
  • CI-friendly API and webhook results

SEO teams

Technical truth at scale

Understand crawling, indexation and internal linking across hundreds of thousands of URLs, with the evidence you need to get engineering time approved.

  • Full link graph and crawl-depth analysis
  • Indexability and canonical resolution
  • Log-style crawl comparisons between runs
  • Exportable evidence for every finding

Agencies

Audit everything, report once

Run scheduled crawls across every client site, track health over time, and produce a report that reads like it took a week.

  • Unlimited projects and client workspaces
  • Scheduled crawls with regression alerts
  • White-label PDF and shareable reports
  • Team seats with per-client permissions

Pricing

Start with a full audit, free.

No trial that hides the findings behind a paywall. Run a complete audit, read every issue, then decide whether you want it running continuously.

Audit

$0

A complete technical audit of one site, so you can see the findings before you commit to anything.

  • 1 project
  • Up to 500 URLs per crawl
  • All 312 technical checks
  • Prioritized issue list
  • Manual crawls
Run a free audit
Most chosen

Team

$79per month

Continuous monitoring for a production site, with the recommendations your engineers need to act.

  • 10 projects
  • Up to 100,000 URLs per crawl
  • Scheduled crawls and regression alerts
  • AI implementation recommendations
  • Crawl-over-crawl comparison
  • 5 team seats
Start free trial

Agency

Custom

Client workspaces, white-label reporting and API access for teams auditing at volume.

  • Unlimited projects
  • Unlimited crawl volume
  • White-label reports
  • Full REST API and webhooks
  • SSO and role-based access
  • Priority support
Talk to us

Questions

What teams ask before their first crawl.

  • How is Metevix different from a standard site crawler?

    Most crawlers stop at detection: they output a list of warnings and leave interpretation to you. Metevix models your whole site — the link graph, template structure and rendered DOM — then scores each finding by its measured impact on your specific pages and gives you the change that resolves it.

  • Does Metevix render JavaScript?

    Yes. Every page is fetched and rendered with JavaScript executed, which is how modern search engines process pages. Metevix reports on both the raw HTML response and the rendered DOM, and flags content that only exists after hydration.

  • Do I need to install anything on my site?

    No. Metevix crawls your site over HTTP the same way a search engine does. There is no script to embed, no build step to change and no code to deploy. For private environments you can allow-list the crawler or supply basic-auth credentials.

  • How long does a crawl take?

    A typical first audit of a few thousand URLs completes in under four minutes. Larger sites scale roughly linearly, and crawl rate is throttled to stay well within what your origin can serve comfortably.

  • Will crawling slow down my website?

    No. Metevix respects robots.txt and crawl-delay directives, adapts its request rate to your server's response times, and backs off automatically if latency rises. You can also set a fixed rate limit per project.

  • Are the AI recommendations safe to apply directly?

    They are written against your actual markup and framework, and each one is checked against the finding that produced it. Treat them as a reviewed pull request from a knowledgeable colleague: read the change, run your tests, then ship it.

Start here

Find out what your site is hiding.

One URL is all it takes. Metevix returns a full technical audit, ranked by impact, in a few minutes.

Free. No credit card, no installation, no changes to your site.