Technical SEO Checklist for Small Business Websites

Technical SEO for a small business is the unglamorous part of being findable. The phone number can be correct, the photos can be yours, and the page can still be invisible because a staging plugin left noindex on, or because the address on the truck redirects three times and dies. This checklist is the list we want a page to survive before anyone spends a month on content or ads.

It is not a score. Google does not publish a technical SEO grade, and a vendor grade out of 100 is only as honest as the rubric behind it. Use the rows as observations. The live version of the same observations is the free website audit. It fetches one public URL and reports the status, the tags, the byte size, and the time. If Google PageSpeed Insights answers, those figures are labeled as Google's. Nothing is filled in when a measurement is missing.

The work this checklist supports is SEO on a site that already says what you do. The build, when the editor cannot hold the fixes, is website development. Speed, which people often mix into this list, has its own article: Core Web Vitals for small business websites.

Start with the URL a customer uses

Write down the address that is printed, advertised, and linked from the Google Business Profile. Check that one, then check the service page you most want a stranger to land on. Homepages get the polish. The water-heater page, the consult page, or the emergency page is the one that has to work.

A plumber in Chicago and a firm reading the Chicago city page have the same technical questions. The city page is market context. It is not a local office, and it does not change what HTTPS means.

HTTPS and the status code

Open the URL and look at the scheme on the final address, not the one you typed. If the site still answers on HTTP, the customer and the crawler are on an unencrypted page. A redirect from HTTP to HTTPS is the right shape. A chain of three or four hops is a delay. Google's site move documentation says to prefer permanent redirects such as 301 or 308, and to keep chains short, ideally no more than three and fewer than five.

The status code has to be a success for a page you want indexed. A 404 on a URL you still print is a broken promise. A 500 is an outage, not an SEO theory. Soft problems hide inside a 200: a page that says "not found" in the heading while the server says OK. Read the words. The checker can only see the code and the HTML.

Certificate errors are a stop. An expired or mismatched certificate is not a title-tag problem. Fix it before you interpret anything else. The audit tool will not ignore a bad certificate in order to scrape the page.

Titles and meta descriptions

Google's title link documentation says every page should have a title element, that the text should be descriptive, and that there is no character limit. The link shown in search is truncated to the device. Google may rewrite a title using the heading, the visible title, or other prominent text when the title element is vague, boilerplate, or out of date.

Write a title a customer can tell apart from the next page on your site. "Home" fails that test. So does a title that repeats the city and the service so many times that it looks stuffed. Brand the homepage. Put the specific service on the service page, with the business name once.

The meta description is a candidate for the snippet, not a guarantee. Google's snippet documentation says snippets are built from page content, and that the meta description is used when it describes the page better than that content. There is no length limit, and the snippet is truncated to the device. Identical descriptions on every URL do not help when several of those URLs appear together. Write the description for that page. If you only have time for a few, do the homepage and the pages that already get visits.

One obvious heading

The H1 should say what the page is. Google's title guidance notes that several headings with the same visual weight make the main title unclear. You do not need a superstition about "only one H1." You need a reader, and a crawler, to see which sentence is the point. A hero that is a picture of text has no heading. The words have to be text.

Canonical and robots

A canonical link is how you name the URL that should survive when copies exist: www and non-www, HTTP and HTTPS, or a parameter a campaign added. The new page after a redesign should point at itself. Pointing the new page at the old URL, or leaving two canonical tags in the head, makes the hint useless.

Robots rules are sharper. Google's robots meta documentation says noindex asks that the page not be shown in search results, and that the more restrictive rule wins when tags conflict. none means noindex and nofollow. A development site that launches with noindex still in a plugin will look finished in a browser and absent in search. Check the meta tag and the X-Robots-Tag header. The audit tool reports both when they are on the response.

Robots.txt is a separate file. If it disallows the URL, a crawler may never see the noindex you set, and Google has said a disallowed URL can still be indexed from other signals, with a title it has to guess. Do not use robots.txt as a substitute for noindex when you need a page kept out.

The viewport meta tag is on this list because a missing viewport is how a phone gets a desktop page squeezed into a narrow window. It does not prove the layout is comfortable. Walk the mobile-friendly checklist on a phone you did not design the page on.

Images and structured data

Count img elements that have no alt attribute. Empty alt is allowed for decorative images. Missing alt is the miss. Google's image guidance treats alt text as a description. A string of cities is not a description of a photograph of a van.

Background images in CSS will not show up in an img count. If the only picture of the work is a background, there is nothing for alt text to describe. Put a real image in the document when the picture is content.

JSON-LD is the structured data this checklist can see without pretending to validate a rich result. A script that does not parse is a bug. A script that parses is not a promise that Google will show a star, a price, or a map. The rules for a local business, and the warning against marking up your own reviews, are in schema markup for a local business.

Open Graph tags matter when the link is pasted into a message. They are a sharing concern. Treat a missing og:image as a preview problem, not as a ranking defect.

What the checklist leaves out on purpose

Backlinks, keyword ranks, and the map pack are not technical properties of one HTML file. The local SEO guide covers the profile, the service area, and why a page per suburb is usually a mistake. How long SEO takes is the timeline conversation, and it is not a date you can buy.

Ads will not repair a noindex tag. Send paid traffic to a URL you have already fetched, with a title that matches the ad and a form a person receives. The beginner Google Ads guide assumes the page exists.

A pass is not a quote

Public ranges for retainers and one-time audits live in the SEO pricing guide. They are market talk, not a CodeHunterz invoice, and they are not a ranking guarantee. Growth, which can include SEO, 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.

If the checklist fails because the theme cannot express a title, a canonical, or a redirect, the conversation is a rebuild, not another plugin. The website cost guide is the budget context. Run the free website checker on the URL you care about, keep the result, and book a call if you want the fixes scoped in writing.