Skip to content
RankX AI

AI Search Visibility Platform

Find what keeps your pages out of AI answers.

RankX AI crawls your site, scores every page, and ranks every finding by what it costs you: start with the free audit, no account and no card needed.

7-day free trial. No credit card required.

app.rankxai.com/site-audit
Website AuditWhat keeps marnwellkitchens.co.uk out of answers, and how bad each one is · Full crawl, 41 pagesSample data

92/ 100 site health

Ranked below by what each one costs the score: one critical finding and four warnings.

FindingsWorst first
  • CriticalThin content
  • WarningNo structured data
  • WarningMissing H1
  • WarningBroken internal links
  • WarningSlow LCP on 4 pages

Every finding opens into the pages it affects, with a fix written for your platform. Marnwell Kitchens is invented; the figures are the sample scenario’s.

What it measures

RankX AI turns a crawl into a score, a ranking and a plan.

A website audit is a crawl that scores your site and ranks what it finds. The terms below decide what you fix first, so RankX AI defines them before showing a number.

  • Health score

    One number from 0 to 100 for the whole crawl. Every critical finding subtracts from it, weighted by how many of your crawled pages it affects.

    Only critical findings move the score. A warning is listed and ranked, never subtracted.

  • Severity

    Findings are graded critical or warning, then ranked by how much each one costs your score, so the list always reads worst first.

    The order is the plan. An alphabetical list treats a broken link like a missing favicon.

  • Pages crawled

    A full crawl covers up to 100, 250, 500 or 750 pages a run, and every page it reaches gets its own score, its own raw signals and its own history.

    A homepage-only check cannot see the page a buyer actually lands on.

  • Fix plan

    RankX AI reads the findings and writes a ranked plan: steps for the platform your site runs on, an effort grade on each fix, and the quick wins pulled out.

    Left as a report, findings get read once. Turned into tasks, they get done.

A check that did not run is reported as not checked, never as a pass, and never moves the score. The figures in the pictures are the sample scenario’s.

What it reads

The crawl reads every page the way a search engine does.

Content, markup, links and performance: each family of checks is a reason pages get skipped, from thin content and missing H1s to broken internal links and slow LCP.

Thin content and pages with nothing to quote

The crawl counts the words on every page and reads the titles, meta descriptions and H1s over them: a page that says too little, or repeats another page’s title, gives an engine nothing to quote.

A page with nothing to quote is a page with nothing to cite.

Structured data missing where it matters

Schema.org markup, Open Graph tags and canonicals are read off every page: what is declared, what is missing, and the canonical chains that end somewhere broken.

An unmarked page makes the engine guess, and it rarely guesses your way.

Dead ends: broken links and orphan pages

Internal links are followed and their status codes recorded, redirect chains traced, and the sitemap compared against the crawl for orphan pages and URLs that no longer resolve.

Every dead end spends crawl budget a live page needed.

Slow pages that cost you readers

Server response time, load time, page weight, compression and render-blocking scripts are measured on every page, with a Lighthouse performance card on the finished audit.

A page that loads slowly gets fetched less and quoted last.

Checks that need a rendered browser, like layout shift, run on single-page scans, and a check that cannot run is reported as not checked rather than passed. The values in the pictures are the sample scenario’s.

What you see

A finished audit is a ranked list, a page score and a comparison.

The audit opens on a health score for the whole site, then the work lives in three views: the finding and its pages, each page and its signals, and the change since the last run.

One finding, every page it affects

A finding opens into the pages behind it, with a fix written for the check and for the platform your site runs on.

Every page carries its own score

Each crawled page gets a score of its own, the raw signals read off it, and a history across runs, so a redesign shows up as a line on a chart rather than a memory.

Two runs, and what changed between

Every audit is diffed against the one before. A check that starts failing is tagged new, one that stops is tagged fixed, and the health score is charted across runs.

A snapshot, not a stateAn audit is your site on the day of the crawl. Sites drift between runs: content edits, plugin updates and redirects all move the score.

Re-runs are scheduledScheduled audits re-check your homepage on the cadence you choose, and a full crawl runs when you ask for one, so drift is caught between full runs.

The diff is the storyOne run says what is wrong. Two runs say whether it is getting fixed, and that direction is the number worth putting in a report.

Sample data with invented brands: RankX AI does not publish measured audit results for companies that never asked.

Start your first full crawl

After the crawl

Findings become a plan, and the plan becomes tasks.

Nothing here ends at a report: RankX AI writes fixes for the platform you run, files them as a tracked queue, and Content Studio rewrites the pages behind the worst findings.

  1. Turn findings into a fix plan

    RankX AI writes the plan from your findings, with steps for the platform your site runs on and an effort grade on every fix.

  2. Create tasks from the plan

    One click files the plan as tracked tasks, deduplicated against work already open, and the queue orders everything by priority and age.

  3. Rewrite the pages in Content Studio

    The pages behind your findings get briefed, drafted and pushed to WordPress as a partial update, with your SEO meta left untouched.

Task creation from a plan is capped and deduplicated against work already open, so re-running an audit never floods the queue.

Questions

What people ask about website audits.

Direct answers about the crawl, the score and the plan, including what an audit will not do to your site.

How many pages does the website audit crawl?

A full crawl covers up to 100, 250, 500 or 750 pages, billed as the smallest tier that covers your site, and your plan sets the per-audit page cap. A single page is checked as its own smaller action rather than billed as a crawl, and scheduled re-checks read the homepage only.

Does running an audit change anything on my site?

No. The crawl requests pages the way a search engine crawler does and reads what comes back: the HTML, the links, the response codes and the timings. It never logs in, never submits a form and never writes anything to your site. Watch your server logs during a run and you will see page requests, nothing else.

How often should I re-run a website audit?

Re-run a full crawl after you ship meaningful changes, because the comparison is where the value is: RankX AI diffs each audit against the previous one and tags every finding as new, fixed or ongoing. Between full crawls, scheduled audits re-check your homepage on the cadence you choose, so drift on your busiest page is caught early.

How do audit findings become tasks?

RankX AI writes a fix plan from the findings, and one click files it as tracked tasks, deduplicated against work already open. The queue orders tasks by priority and age, so the most severe, longest-waiting fix sits at the top, and every task records which audit run it came from.

Is this the same as the free audit on this site?

The free audit is account-less: give it a domain and it samples your key pages and scores how ready they are for AI search, before any signup. The Website Audit inside the product is the full crawl this page describes: tiered page counts, a score and history for every page, scheduled re-checks, an audit-to-audit comparison, and a fix plan that becomes tasks. They are different tools with one entrance: start with the free one, and the crawl is what the trial adds.

Do the audit checks cover AI readiness?

They are related and deliberately separate. The Website Audit measures your site’s technical condition; AI Readiness measures whether assistants can reach your site and tell what your business does. The two are never combined into one number, because a blend would let a healthy site hide an invisible one. The free audit is where the AI-readiness angle gets checked first.

More detail in the documentation.

Start here

See where you show up in AI answers today.

Add your site and RankX AI suggests the prompts to track, monitors the keywords that matter and audits your pages, with your first results minutes after you finish setup.

Start Free Trial

7-day free trial. No credit card required. Cancel anytime.