Free SEO + AI Audit documentation
The Free SEO + AI Audit crawls your website and shows how Google and AI assistants like ChatGPT, Claude and Perplexity see it. This guide explains every option, how the score is calculated and what to do with your report.
What it does
The Free SEO + AI Audit reads your website the way Google and AI assistants do, then gives you one score, six category scores and a list of problems ranked by impact. Every problem comes with a plain-English reason why it matters and a clear fix.
It looks at six areas:
- Technical SEO: status codes, redirects, HTTPS, robots.txt, XML sitemaps, canonical tags, noindex, mobile viewport and mixed content.
- On-page SEO: titles, meta descriptions, duplicates, H1 headings, image alt text and social sharing tags.
- Content quality: thin pages, internal links, subheadings and whether your homepage copy reads as AI-written.
- Speed and weight: server response time, HTML size, compression and script count.
- AI search readiness: which AI crawlers your robots.txt allows, JavaScript-only content, structured data, FAQs and llms.txt.
- Trust and authority: About, Contact and privacy links, author bylines, dates and a favicon, the E-E-A-T signals people and search engines look for.
The crawler, RVAuditBot, is built in-house. It does not use outside SEO services, it never logs in or submits forms, and it does not change anything on your site.
Quick start
- Open the audit page and create a free account (name, email and a password of at least 10 characters), or log in if you already have one.
- Your site: type your website address, for example
yourwebsite.com, and pick the kind of business it is. - Your goals: choose one or more goals, such as more leads or getting cited by AI search.
- Your plans: optionally add your budget, timeline, biggest challenge and competitors, then press Run my free audit.
- Watch the progress bar as the crawler reads robots.txt and sitemaps, crawls pages, checks AI search readiness and writes your report.
- Read your report on the same page. Start with the list titled Fix these, in this order.
You do not need to type https://. If you leave it out, the tool adds it for you.
Every option explained
The audit is set up in a short three-step form. Only the website, business type and at least one goal are required.
| Field | What it does | Required? |
|---|---|---|
| Website to audit | The address the crawl starts from. Only http:// and https:// addresses on public websites can be audited. | Yes |
| Business type | One of 7 choices: Local business, E-commerce, SaaS / startup, Agency / services, Personal brand, Blog / publisher or Something else. It is saved with your report for context. | Yes |
| Goals | Pick any of the 7 goals. Each goal adds a "Based on your goals" card to the report with the categories to focus on and up to 3 issues to fix first. | At least one |
| Monthly marketing budget | One of 5 ranges, or "Not sure yet". Saved for context. | No |
| Timeline | ASAP, This month, Next 3 months or Just exploring. Saved for context. | No |
| Biggest challenge | Free text, up to 1,000 characters. | No |
| Main competitors | Free text, up to 300 characters. These sites are not crawled; the note simply helps a follow-up conversation. | No |
How each goal shapes your priorities
| Goal | Categories it focuses on |
|---|---|
| More organic traffic | On-page SEO, Content quality, Technical SEO |
| More leads and enquiries | Trust and authority, On-page SEO, Speed and weight |
| Rank in local search and Maps | Trust and authority, On-page SEO, Technical SEO |
| Get cited by ChatGPT and AI search | AI search readiness, Content quality, Trust and authority |
| More online sales | Speed and weight, Technical SEO, On-page SEO |
| Fix technical problems | Technical SEO, Speed and weight |
| Build brand authority | Trust and authority, Content quality, AI search readiness |
Goals change which fixes are highlighted first. They do not change the scores, so two people auditing the same site at the same moment get the same numbers.
Reading the results
Overall score and grade
At the top you see a score out of 100, a grade, a count of issues at each severity, and how many pages were crawled, how many URLs were discovered and how many checks passed.
| Overall score | Grade |
|---|---|
| 90 to 100 | Excellent |
| 75 to 89 | Good |
| 55 to 74 | Needs work |
| 0 to 54 | Poor |
Category scores
Six cards show each category's score, its number of issues and its number of passed checks. Click a card to filter the issue list to that category, and click it again (or Show all issues) to go back.
AI search readiness panel
This panel shows whether your content is in the HTML, whether llms.txt is published, which structured data types are on your homepage, and an AI-written likelihood score for your homepage copy (0 to 100) from the same engine as the AI Website Detector. Below that is a row for each of 14 crawlers marked Allowed or Blocked: Googlebot and Bingbot, plus 12 AI crawlers from OpenAI, Anthropic, Perplexity, Google, Apple, Common Crawl, Meta and Amazon.
The issue list
Issues are sorted by severity (Critical, High, Medium, Low), and within each severity by how many pages are affected. Open an issue to see:
- Why it matters in one or two sentences.
- How to fix it in practical steps.
- Affected pages, showing up to 8 example URLs, or "Site-wide" for problems that apply to the whole site.
What is already working and pages crawled
The What's already working list shows every check you passed. The Pages crawled table lists each page with its HTTP status, word count, server response time and number of issues. Response times over 800 ms are highlighted as a warning and over 1,500 ms as slow.
How it works
1. Site-level probes
RVAuditBot first reads your robots.txt and obeys it for its own user agent. It finds your sitemaps from the Sitemap: lines in robots.txt, or tries /sitemap.xml and /sitemap_index.xml. It also checks for /llms.txt, tests whether http:// redirects to https://, and requests a made-up URL to confirm missing pages return a 404 status.
2. The crawl
Starting from your homepage and sitemap URLs, it crawls up to 25 pages within 90 seconds, following internal links. Your domain with and without www. counts as the same site. It skips files such as PDFs, images, scripts and documents, and paths like cart, checkout, login and admin areas.
3. The checks
Each page is checked for the items in the table below. A few checks are site-wide, such as robots.txt, sitemaps and the AI crawler rules, which are tested against your homepage URL.
| Check | Threshold used |
|---|---|
| Title length | Flagged if longer than 60 or shorter than 20 characters |
| Meta description length | Flagged outside 70 to 160 characters |
| Very thin / thin pages | Under 100 words / 100 to 299 words |
| Few internal links | Fewer than 3 internal links |
| Long pages without subheadings | Over 600 words and no H2 |
| Few lists or tables | Over 500 words, fewer than 2 lists and no tables |
| Sluggish / slow server response | Over 800 ms / over 1,500 ms to first byte |
| Heavy HTML | Over 300 KB |
| Many external scripts | More than 25 |
| Redirect chains | 2 or more hops |
| Hard-to-read URLs | Uppercase letters, underscores or over 115 characters |
| Content needs JavaScript | Under 60 words of text in an app-style page shell |
| Copy reads as AI-generated | Homepage AI-written likelihood of 65 or more |
4. Scoring
Every category starts at 100 and loses points for each issue it contains:
| Severity | Full penalty |
|---|---|
| Critical | 30 points |
| High | 15 points |
| Medium | 8 points |
| Low | 3 points |
Site-wide issues take the full penalty. Page-level issues are scaled by how much of the site they touch: the penalty is multiplied by 0.45 + 0.55 × (affected pages / pages crawled). So one bad page out of 25 costs a little under half the full penalty, while a problem on every page costs all of it. A category never drops below 0.
The overall score is a weighted average of the six categories:
| Category | Weight |
|---|---|
| AI search readiness | 23% |
| Technical SEO | 20% |
| On-page SEO | 20% |
| Trust and authority (E-E-A-T) | 15% |
| Content quality | 12% |
| Speed and weight | 10% |
Tips for the best results
- Enter your main domain. The crawl starts from the address you type and stays on that domain, so use the version your visitors see.
- Fix Critical and High issues first. They cost the most points and usually have the biggest effect on traffic.
- Check your robots.txt rules for AI crawlers. Blocking AI answer bots (such as OAI-SearchBot, Claude-SearchBot or PerplexityBot) is a High issue. Blocking training-only bots is a Low issue, because that is a valid choice.
- Keep a sitemap listed in robots.txt. The crawler uses it to find your important pages quickly.
- Make sure your text is in the HTML. If your site is a JavaScript app, server rendering or pre-rendering helps both crawlers and AI assistants read it.
- Run a fresh audit after a round of fixes. Your earlier reports stay in your account, so you can compare scores.
- Save a PDF to share. Press Save as PDF to open your browser's print dialog and save the report for your developer or team.
Limits, privacy and fair use
- Account required. You need a free account to run an audit and view reports. No credit card is asked for.
- 3 free reports per account. This is a lifetime allowance, shown as "reports left" on the form. Audits that fail do not count. When all 3 are used, your past reports stay available and you can book a call for a deeper audit with my team at Thetechhacker.
- Crawl size. Up to 25 pages and 90 seconds per audit. Bigger sites get a sample, not a full crawl.
- Shared capacity. Up to 2 audits run at the same time across the whole site. If both slots are busy, you see "Waiting for a free crawler" until one opens.
- Public sites only. Local, private and internal addresses cannot be audited, and the crawler does not sign in to protected areas.
- Your reports are private. A report link only opens for the account that ran it. To share results, save the report as a PDF.
- Your details. Your account details and goals are used to deliver your report and, if you would like, to follow up about working together. They are never sold or shared. The optional tips email can be switched off at sign-up.
- Respectful crawling. RVAuditBot identifies itself in its user agent and obeys your robots.txt.
Troubleshooting
- "Pick at least one goal so the report can focus on what matters to you." Go back to step 2 and choose a goal.
- "Please enter a website address." or "That does not look like a valid website address." Type a full domain such as
yourwebsite.com, without spaces. - "That domain could not be resolved." Check the spelling and that the domain is live.
- "Internal or local addresses cannot be analysed." Staging and local sites cannot be audited. Publish the site, or audit the live version.
- "We couldn't read any pages on this site (...)." None of the pages returned readable HTML. Make sure the site is public, not behind a password or firewall, and loads in a normal browser. The reason in brackets (such as an HTTP status) tells you where to look.
- "Interrupted by a server restart. Please run it again." Start the audit again. It does not use up a free report.
- "Something went wrong while auditing this site. Please try again." Try again in a few minutes. Failed audits never count toward your 3 free reports.
- "You've used all 3 free reports on this account." Your old reports are still in "Your reports". Book a call to go further.
- "An account with this email already exists. Try logging in instead." Switch to the Log in tab.
- "Use at least 10 characters." Choose a longer password. A short phrase works well. Very common passwords and ones based on your name or email are also declined.
- "Report not found." Report links only open for the account that created them. Log in with that account.
- Fewer pages than expected. The crawl stops at 25 pages or 90 seconds, skips paths blocked in robots.txt, and only follows links on your own domain.
- "Content needs JavaScript" on a site that looks fine. Your text is added by JavaScript after the page loads. Most AI crawlers do not run JavaScript, so server rendering is the fix.