We Built a Free Website Grader. Here Is What It Checks and Why.

We built a tool that scans any website and scores it across four categories: SEO and Visibility, Technical quality, Trust and Authority, and Content Quality. It checks things most free graders skip. It is free, it does not require a login, and it shows you the specific findings behind every score. This post explains what it checks, why those things matter, and how the scoring works.

Why Another Website Grader

There are plenty of website grading tools already. HubSpot's Website Grader is probably the most well known. It checks performance, SEO, mobile readiness, and security. It has been around since 2008 and it does a solid job at what it covers.

But most free graders, including HubSpot's, only look at the homepage. They do not fetch your robots.txt to see if you are accidentally blocking search engines. They do not check whether your sitemap.xml exists, how many URLs it contains, or whether the lastmod dates are current. They do not crawl your internal pages to see if the SEO quality on your about page matches the quality on your homepage. And they rarely check for specific structured data types, the markup that tells search engines and AI systems exactly what your business is and what services you offer.

We built this grader because we needed something deeper for our own work. After scanning 200 local business websites for competitive analysis, we realized that single-page homepage checks miss the problems that actually matter. A site can have a perfect homepage and completely broken internal pages. A site can look great in the browser but have no sitemap, no robots.txt, and no structured data. Those are the gaps that cost businesses visibility in search results and AI citations.

The Four Categories

The grader scores your site across four categories, each worth 10 points. The four scores are added together and scaled to 100. Every deduction comes with a specific finding that tells you exactly what was detected and whether it passed, raised a warning, or failed outright.

What Each Check Actually Means

A score without context is just a number. Here is what the key checks are looking for and why they matter.

Sitemap and Robots.txt

A sitemap is an XML file that lists every page on your site along with when each was last updated. It is how you tell Google, Bing, and AI crawlers exactly what content exists and what has changed recently. Without one, search engines have to discover your pages by following links, which means pages that are not well-linked internally might never get indexed.

A robots.txt file tells search engine crawlers which parts of your site they are allowed to access. It is also where you can explicitly invite AI crawlers like GPTBot, ClaudeBot, and PerplexityBot to index your content. Without a robots.txt, crawlers will still index your site, but you lose control over the process. We check whether the file exists, whether it references a sitemap, and whether it is accidentally blocking Googlebot or all crawlers.

Most website graders skip both of these checks entirely because fetching them requires separate HTTP requests to files that return plain text or XML, not HTML. Our grader makes those requests.

Structured Data Types

Structured data is JSON-LD code embedded in your page that tells search engines and AI systems exactly what your business is, where it operates, what it offers, and how to contact you. The grader does not just check whether structured data exists. It extracts the specific schema types present and evaluates whether they include business-relevant types like LocalBusiness, Organization, ProfessionalService, or Restaurant.

A site with generic WebPage schema gets partial credit. A site with LocalBusiness or ProfessionalService schema gets full credit. A site with nothing gets zero. This distinction matters because AI search engines like ChatGPT and Perplexity rely heavily on structured data to decide which businesses to cite in their responses. We covered this in detail in our guide to AI search optimization for small businesses.

Cross-Page Consistency

This is the check that requires crawling internal pages, and it is the one that most graders cannot do. The grader discovers links on your homepage, selects two to three internal pages (prioritizing pages like about, services, and contact), fetches them, and runs the same analysis on each one.

It then checks whether those internal pages have meta descriptions and proper H1 heading tags. If your homepage is well-optimized but your about page has no meta description, the grader catches that. A site that is consistent across all pages scores higher than one where the homepage is polished but the inner pages were neglected.

Call-to-Action Detection

Some graders penalize sites for not having a contact form on the homepage. That always struck us as arbitrary. A business that uses phone calls and email as its primary contact method should not lose points because it does not have a form element in its HTML. Our grader looks for any conversion pathway: a phone link, an email link, or a contact form. If any of those exist, the check passes. Only sites with no way for a visitor to take action get flagged.

Template Builder Detection

The grader scans the HTML source for signatures of template platforms like WordPress, Wix, GoDaddy, Shopify, and others. This is not an automatic penalty. Template builders are tools, and some businesses use them well. But in the Trust and Authority category, a custom-built site gets full marks while a template site does not. The reasoning is practical: custom-built sites are more likely to have been intentionally structured for search engines and AI visibility, while template sites often ship with default configurations that the business owner never changed.

How the Scan Works

When you enter a URL, the grader runs through five phases in roughly 10 to 15 seconds:

1. Homepage fetch. The server sends a request with full Chrome browser headers to mimic a real visitor. If the site blocks this request (common with Cloudflare-protected sites), it falls back to a headless Chrome browser via Puppeteer to render the page.
2. robots.txt and sitemap. Separate requests fetch these files in parallel. The sitemap URL is extracted from robots.txt if present, with fallbacks to standard locations.
3. Internal page crawl. Links are extracted from the homepage HTML. The grader selects two to three internal pages, prioritizing pages that are likely to be important (about, services, contact, portfolio), and fetches each one.
4. HTML analysis. Every fetched page is parsed with a full DOM parser. The grader extracts headings, images with and without alt text, Open Graph tags, canonical tags, schema types from JSON-LD blocks, word count (excluding navigation and footer boilerplate), and internal link counts.
5. Scoring. Each check contributes a specific number of points to its category. Points are clamped to 10 per category, summed, and scaled to 100.

Results are cached for 24 hours by domain. If someone else scans the same site within that window, they get the cached result instantly. The cache exists to be respectful to the sites being scanned, not to limit usage.

What the Scores Mean in Practice

After testing the grader on dozens of sites, here is a rough guide to what the overall score ranges tend to look like:

85 to 100: Well-built site with intentional SEO structure. Has a sitemap, robots.txt, structured data, proper headings across multiple pages, and current content. These are typically custom-built sites or professionally maintained WordPress installations.

60 to 84: Functional site with gaps. Common pattern: good homepage, but missing sitemap, no structured data, one or two internal pages without meta descriptions. Many WordPress sites with quality themes land here.

40 to 59: Template site with minimal customization. Often missing multiple signals: no structured data, no analytics, outdated copyright, thin content. The site exists but is not working hard for the business.

Below 40: Significant problems. Missing SSL, not mobile-friendly, no sitemap, no structured data, very thin content. These sites may be actively hurting the business by signaling neglect to search engines and potential customers.

The score is not a judgment on the business. It is a measurement of how well the website communicates with search engines and AI systems. A five-star plumber with a 35-rated website is still a great plumber. The website is just failing to represent the business accurately online.

Try It

The grader is live and free to use. Enter any URL and get your scores in about 15 seconds. No account required. No credit card. The four category scores and top findings are shown immediately. If you want the full detailed report emailed to you, enter your email address after the scan completes.

If your score surfaces something you were not aware of, that is the point. Every finding in the report is something you can either fix yourself or bring to your web developer with a specific request. "Add a sitemap" is more actionable than "improve your SEO."

And if you want to see how your competitors are doing, the competitive analysis tool on our Solutions page scans up to 20 businesses in your industry and city and generates an AI-powered prospecting brief. Different tool, different purpose, same underlying infrastructure.

What We Plan to Add

The current version is a medium-depth scan. It is deeper than most free tools but it is not a full technical audit. Some things we are considering for future versions:

• Page speed metrics. Real Core Web Vitals measurement using Lighthouse or the Chrome UX Report API. Currently, we proxy page size as a speed indicator, but actual load timing would be more useful.
• Deeper internal crawl. Scanning 5 to 10 pages instead of 2 to 3, with detection of orphan pages and broken internal links.
• AI visibility scoring. Checking whether the site is cited by ChatGPT, Perplexity, or Google AI Overviews for relevant queries. This would tie the grader directly to generative search performance.
• Competitor comparison. Grade your site alongside two or three competitors and see a side-by-side breakdown.

These are possibilities, not promises. The current tool does what it does well, and we would rather ship something useful and improve it than wait for a feature list that never ships.