Free website audit.

What a free website audit is for

A website checker is useful when you want a straight reading of one page before you buy a rebuild, an SEO retainer, or another month of ads. Owners usually meet the page the way a customer does: on a phone, from a map listing, or from an ad. The HTML behind that visit is still worth reading. A missing title, a certificate error, or a noindex tag left on after a redesign will not announce itself in the design file.

This page is the checker. It is also the explanation of how to read the result. The tools hub is the short index. If you want the longer walkthrough of the same checks, read the technical SEO checklist. If you are deciding whether the site should be replaced, the symptoms are in signs your website is losing customers. The build, when you want one, is website development. Search work that continues after the page is fixed is SEO and, for a business that depends on a map listing, local SEO.

The checker looks at one URL you submit. It does not crawl the rest of the site, log into a CMS, or open your Google Business Profile. A homepage can look tidy while the service page people actually land on is the one with the problem. Run the important URL, not only the root.

Results come from the request, not from a formula

Every row is labeled Pass, Check, Fail, or Note. Pass means the thing we looked for was present and did not trip a specific problem, such as a noindex rule. Check means you should look, because something a visitor or a crawler might need was missing or repeated. Fail means the response itself was a problem: the certificate did not verify, the final URL was still HTTP, or the status code was an error. Note means we measured something and are not grading it. Response time and page size are notes. There is no overall number at the top that we calculated.

That distinction matters. Plenty of "free website audit" pages print a score out of 100 from a rubric the vendor never publishes. A 73 can mean anything. Here, a title row says how many characters were in the title element and quotes the text. An image row says how many img elements had an alt attribute. If a section cannot be measured, the row says why. We would rather show a gap than fill it.

Google's own title link documentation says there is no character limit on a title element, and that the link shown in search is truncated to fit the device. The checker therefore does not fail a title for being "too long." It tells you the length and the words. You can decide whether a customer can tell the page apart. The same idea applies to the meta description. Google's snippet documentation says snippets are created from page content, and that the meta description is used when it describes the page better than that content. A missing description is marked Check, not as a penalty we invented.

HTTPS, status codes, and redirects

The first questions are whether the connection is trustworthy and whether the URL you typed is the URL that answered. The checker follows up to five redirects, and only when each target is still a public http or https address on port 80 or 443. A redirect toward a private network, localhost, or a non-web scheme is not followed. The result stops on the last public response and says so.

A final URL on HTTPS is a pass. A final URL that is still HTTP is a fail, because the page the customer would read was not encrypted. If you typed HTTP and the site sent you to HTTPS, the row says that the hop happened. Google's site move documentation treats an HTTP to HTTPS change as a move that deserves permanent redirects and a clean canonical, not a chain of temporary hops.

Status codes are reported as the server sent them. A 200 is a pass. A 404 or a 500 is a fail, and the on-page rows still appear when the error page was HTML, because an error document can carry a noindex tag or a title that says the wrong thing. A long redirect chain is marked Check once it passes three hops. Google advises keeping chains short, ideally no more than three and fewer than five, because each hop adds delay and some clients will not follow a long chain. The list in the result is the chain we actually walked, with the status on each hop.

Certificate failures are reported on their own. If the certificate is expired, self-signed, or does not match the host name, we do not turn certificate checking off to "see the page anyway." The body is not read, and PageSpeed Insights is not called. Fix the certificate, then run the check again.

Title, description, and the H1

The title element is the strongest hint you give Google for the blue link, and Google may still rewrite it. The documentation lists the sources it considers: the title element, the visible title, headings, Open Graph, and other prominent text. A title that says "Home" on a plumbing company in a specific city gives Google very little to use, and it gives the customer even less. The checker quotes the title it found and counts the characters. More than one title element is marked Check, because the page is then arguing with itself.

The meta description is the pitch you hope will be used as the snippet. Write it for the page, not for the whole company. A homepage can carry the business. A water-heater page should say water heater. Google also says identical descriptions across a site are not helpful when several of those pages show up together. This tool sees one URL, so it cannot tell you that every page shares the same sentence. If the homepage description could be pasted onto any of your competitors, rewrite it before you worry about a plugin.

The H1 is the heading in the document, which is not always the largest text a person sees if the design hides it. Zero H1s is a Check. One or more is a pass, and the count is in the row. HTML allows more than one. Google's title guidance says that when several headings have the same visual weight, it may not be clear which text is the main title. Use the count as a prompt to look, not as a rule that a second H1 is a defect.

Canonical, robots, and the viewport

A canonical link tells crawlers which URL is the one you want treated as the original when copies exist. The checker reports the href it found, resolved against the final URL when the href was relative. No canonical is a Check. Two canonicals are a Check. A canonical that points at a different host is still reported honestly. You have to decide whether that host is really you. On a redesign, the new page should canonical to itself, not to the URL you just retired. The migration notes are in the website migration checklist.

The robots meta tag and the X-Robots-Tag header are how a page asks not to be indexed. Google's robots meta documentation says noindex means the page should not be shown in search results, and that a conflicting set of rules follows the more restrictive one. A staging site that ships with noindex still in the theme is a quiet way to launch a site nobody can find. The checker marks noindex or none as Check and quotes the tag or header. Absence of a robots meta tag is a Note: the tag is not what is blocking indexing. Robots.txt is a different file, and this tool does not fetch it. A disallow in robots.txt can stop a crawler from seeing the noindex you thought you set.

The viewport meta tag is the page telling the browser how to size itself on a phone. Missing content is a Check. The row quotes the content when it is there. A viewport tag does not prove the layout fits a thumb. For that, use a phone. The mobile-friendly checklist is the physical version of this test: tap to call, readable type, and a form that can be submitted.

Images, Open Graph, and structured data

Alt coverage is counted from img elements in the HTML we downloaded. An alt attribute that is present but empty is counted as present, because an empty alt is the correct mark for a decorative image. A missing attribute is the gap. The row also says how many of those alts had text. Pictures inserted by JavaScript after load are not in this count. If your hero is a background image in CSS, it will not appear as an img, and it also will not have an alt. Google's image SEO guidance treats alt text as a description of the image for people who cannot see it and as a clue to what the image shows. Do not stuff it with a city name. Describe the photograph.

Open Graph tags (og:title, og:description, og:image) control how a link looks when someone pastes it into a message or a social network. They are not a ranking checklist. The checker marks all three present as a pass and anything less as a Note, and it names which ones it found. A missing image is the one people notice, because the preview becomes a blank box. Use a real photo of the work, not a logo stretched into a landscape frame, unless the logo is actually what you want shared.

Structured data is reported as JSON-LD only. If a script with type application/ld+json is present, we try to parse it and list the @type values, including types inside @graph. A script that does not parse is a Check. No JSON-LD is a Note. We do not claim a rich result, and we do not run Google's Rich Results Test for you. Google's structured data introduction says structured data helps Google understand a page, and the local business documentation says Google does not guarantee that a feature will show. Eligibility is stricter than "some JSON exists." For a business with a published address, Google's LocalBusiness documentation lists name and address as required properties. Do not invent a street address to satisfy the field. The longer explanation, including why review markup is the wrong place for your own testimonials, is the schema markup guide.

Page size, load time, and PageSpeed Insights

The byte count is the body we received. If the server compressed it, the row shows both the wire size and the decoded size. The clock is the time from our request until the headers of the final response, and until the download finished. That is a server in a data center talking to your host. It is not your customer's phone, and it is not Core Web Vitals. Largest Contentful Paint, Interaction to Next Paint, and Cumulative Layout Shift describe real visits. A fast HTML download can still wait on a huge photograph. The field thresholds and what they feel like are in Core Web Vitals for small business websites.

When the check can, it also calls the public PageSpeed Insights API for a mobile run. If Google returns a Lighthouse performance score, we show that number and label it as Google's lab score. Lighthouse display values, such as a Largest Contentful Paint string, are shown as Google wrote them. If the response includes Chrome UX Report field data, we show the percentile and the category Google used (FAST, AVERAGE, or SLOW). If Google says the figures are an origin fallback, we say that they describe the whole site rather than the exact URL. If the API is slow, rate limited, or empty, the section says it did not return data. We do not estimate a score to fill the space, and we do not average a lab number with a field number.

Google notes that lab data is a simulated load and field data is a 28-day history of real Chrome users, so the two can disagree. A single lab run also moves from one attempt to the next. Treat a Lighthouse figure as a clue about this fetch, and treat field data as the experience Google has already observed. Neither one is a promise that a redesign will rank.

What this website checker will not tell you

It will not tell you your position for a keyword. It will not count backlinks, referring domains, or toxic anchors. It will not audit the Google Business Profile, the citations, or the reviews. It will not read your ad account or tell you which search terms wasted money. It will not submit the form on the site and confirm that a person received it. It will not render JavaScript the way a browser does, so a title injected only after a script runs can be invisible to this fetch. Say that out loud before you treat a pass as a clean bill of health.

It also will not price the repair. Market ranges for a site, a local SEO retainer, and the other common budgets are in the pricing guides, including website cost and SEO pricing. Those pages are public ranges, not a CodeHunterz invoice. The Website plan starts at $6,000 one time after a 20% launch offer (regular starting price $7,500). Growth starts at $2,400 per month after a 20% launch offer (regular starting price $3,000 per month). Ad spend is separate when ads are included. Limited-time launch offer, final quote after free strategy call.

Local pages on this site describe the market, not a storefront we claim to keep. A Chicago service company can read the Chicago page and a plumbing company can read the plumber page for how we scope the work. The checker itself is the same in every city. It does not boost a URL because of a place name.

What to do with the result

Fix the fails first. A certificate error, an HTTP final URL, or a 404 on the address printed on the truck is more urgent than an Open Graph image. Then read the Check rows. A noindex tag, a missing viewport, a missing title, or a pile of images without alt attributes are ordinary repairs. Notes are context. A two-second download of a small HTML file can still hide a slow hero. Open the page on a phone after you read the numbers.

If you want a person to look at the same URL with you, book a free call. Bring the result, or run it during the call. You should leave knowing whether the next step is a repair on the current site, a redesign, or search and ads work on pages that already hold up. The written audit form on the homepage is a different offer: you send the URL and we reply with a plan. The tool on this page is the measurement you can see immediately. Neither one reserves a project, and neither one promises rankings, calls, or a full calendar.

Use the checker on a page you are allowed to request. Do not point it at someone else's private admin, a link-local address, or a network that is not yours. We rate limit repeats so the tool stays a check rather than a crawler. One page, a few times in ten minutes, is the shape it was built for.