The short answer: technical SEO means making sure search engines can find, read and understand your pages. For a new website the work is mostly decided before launch: a clean structure, the right pages indexable, a sitemap, sensible titles, fast pages and tracking in place. Fixing these later costs more and loses weeks of visibility.
A beautiful site that cannot be crawled is invisible. The good news is that the basics are a checklist, not a mystery.
1. Make sure the site can be crawled
- Remove any "noindex" left from staging. This is the most common launch mistake. A development site often blocks search engines, and the setting travels with the files.
- Check robots.txt. It should allow the pages you want found and point to your sitemap.
- Publish an XML sitemap listing every page you want indexed, using the final addresses.
- Serve the site on one address. Choose https and one version of the domain, and redirect the others to it.
2. Plan the structure and the addresses
- Short, readable addresses. Words people can understand, in lower case, separated by hyphens.
- A logical hierarchy. Main pages, then services, then articles, with every page reachable from another.
- One page per topic. Two pages answering the same query compete with each other.
- Internal links with clear anchor text. They tell search engines what each page is about.
3. Get each page's basics right
- A unique title and description for every page, with the main phrase near the start of the title.
- One main heading that says what the page is.
- A canonical tag pointing at the preferred address, so duplicates do not split the signal.
- Descriptive image names and alternative text for images that carry meaning.
4. Speed and mobile
Most visitors arrive on a phone. Test the pages on a real phone on a normal connection, not only on a fast computer. Compress images, load only the scripts you need and avoid layout shifts. Our guide to Core Web Vitals explains the measures in plain language.
5. Structured data
Structured data describes the page to machines: the organisation, an article, a frequently asked question, a breadcrumb. Use the types that match what is really on the page, and keep the marked-up text identical to what a visitor sees. Test it with a validation tool before you publish.
6. Tracking and search tools
- Install analytics and confirm it records visits.
- Verify the site in the search engine's webmaster tool and submit the sitemap.
- Define the key action, such as a form submission or a message, as a goal.
Do this before launch, so you can see what happens in the first days.
7. After launch
- Check that the important pages are being indexed, and fix any that are not.
- Look at errors reported in the search tool: missing pages, blocked pages.
- Re-run the checklist after any redesign or move. Redesigns are where rankings are lost.
An illustration: a launch day check
Imagine a new site is about to go live. The team opens the home page and views the source: the robots setting still says noindex, left over from development. They remove it. They open the sitemap and find that it lists the staging addresses, not the real ones. They regenerate it. They test a form and confirm the enquiry arrives. They open the site on a phone and find a very large image slowing the first screen, and they compress it.
None of this is difficult. All of it can cost weeks of visibility if it is found after launch instead of before.
Monthly maintenance
- Check the search tool for new errors and pages that dropped out of the index.
- Look for broken links, especially after content changes.
- Re-test speed on a phone after any new plugin, script or large image.
- Update titles and descriptions on pages that get impressions but few clicks.
An hour a month prevents most surprises.
Free tools that cover most of this
- The search engine's webmaster tool for indexing, sitemaps, errors and real-visitor speed data.
- A public page speed test for quick checks on a phone profile.
- A structured data testing tool for the markup.
- A crawler that follows your links and lists broken ones.
- Your browser's developer view for checking titles, headings and robots settings.
A small business can do a solid technical check with these and an afternoon.
How search engines find and read a page
Understanding the basic path helps you fix problems faster. A search engine first discovers a page, usually by following links or reading a sitemap. It then fetches the page, reads the content and the instructions in the code, and decides whether to store it. Only stored pages can appear in results.
Problems can occur at each step. A page with no links pointing to it may never be found. A page blocked in the robots file may not be fetched. A page that depends on scripts that fail to load may be read as nearly empty. A page marked with a "noindex" instruction will be fetched but not stored. When a page is missing from results, ask which step failed, and use the inspection tool in the search console to see how the engine sees it.
Set the rules at template level
On most sites, technical settings are applied through templates, not page by page. That is good news: a rule written once applies everywhere. It also means one mistake in a template can affect hundreds of pages.
- Titles and descriptions. Build fields that editors can fill, with sensible defaults, instead of repeating one fixed title everywhere.
- Canonical tags. Make sure each page points to its preferred address, especially where filters or tracking parameters create duplicates.
- Headings. Use one main heading per page and a logical order below it.
- Images. Require descriptive text and sensible dimensions for each image.
- Social sharing data. Include the fields that control how a page looks when shared.
Test each template with a real page before launch, and again after any theme or plugin update.
Launch mistakes that are easy to avoid
- Leaving the site hidden from search after launch because a setting used during development was never turned off.
- Launching without redirects from old addresses when a site replaces an earlier one.
- Blocking the wrong files in the robots file, such as the stylesheets and scripts needed to display the page.
- Forgetting the sitemap or listing addresses that redirect or return errors.
- Skipping mobile testing on real phones.
- Adding tracking late, so the first weeks of data are lost.
A short written checklist that someone signs off on before launch prevents most of these. Keep it with the project, and update it each time something new goes wrong.
How we handle it
We run this checklist on every site we build or manage, because it is cheaper to prevent than to repair. If you are about to launch or already have a site that is not being found, talk to us about an audit.