A useful new website needs more than a keyword in every title. Search engines must be able to fetch the pages, understand how they relate, and find content that actually resolves a reader's task. This checklist puts those dependencies in launch order so a team can fix blockers before polishing minor details.
Start here: make the site accessible, indexable, coherent, and useful. Search optimization cannot compensate for a page that is blocked, duplicated, empty, misleading, or difficult to use.
Phase 1: confirm that the site can be crawled
- Use HTTPS everywhere. Redirect each HTTP URL to its HTTPS equivalent and avoid mixed-content requests.
- Return the right status codes. Published pages should return 200, moved pages should redirect once to the closest replacement, and removed pages without a substitute should return 404 or 410.
- Review robots.txt. Make sure it does not block pages, CSS, JavaScript, or images that search engines need to render public content.
- Remove accidental noindex rules. Check templates, staging settings, response headers, and page-level meta tags.
- Provide crawlable links. Important destinations should be reachable through ordinary HTML anchor elements, not only through a search box or scripted interaction.
Google's search guide for developers specifically recommends crawlable links, descriptive metadata, semantic HTML, accessible resources, and a sitemap. Test the final production URLs, not just a preview environment.
Phase 2: give every useful page one clear home
- Choose one preferred URL per piece of content. Redirect duplicate protocol, hostname, slash, and legacy variants where practical.
- Add a self-referencing canonical. The canonical should name the live preferred URL, not a staging address or a different article.
- Use descriptive paths. A person should be able to infer the subject from a URL such as
/guides/website-speed-optimization/. - Keep taxonomies intentional. Do not publish empty tags, duplicate date archives, or categories created only to repeat a phrase.
- Map redirects before launch. If an old site exists, pair valuable old URLs with their closest new destination instead of sending everything to the homepage.
The Google SEO Starter Guide explains that duplicate URLs are not automatically a spam violation, but they can confuse users and waste crawling effort. A simple architecture is easier for people to navigate and easier to maintain.
Phase 3: make the search result accurate
- Write a unique title. State the page's actual task or subject. Keep the visible H1 and title element aligned rather than using a sensational title that the page does not fulfill.
- Write a useful meta description. Summarize what the reader can accomplish and what the page includes. Do not duplicate the same boilerplate across the site.
- Use one clear H1. Subsections should follow a logical H2 and H3 hierarchy.
- Set Open Graph metadata. Use an accurate title, description, URL, and image when a page has a suitable share image.
- Check the visible first paragraph. It should answer what the page is for without forcing the reader through a long preamble.
Google may generate title links and snippets from several page signals, not only the title and description tags. Its title-link guidance recommends concise, descriptive, page-specific title text and consistency among prominent headings and metadata.
Phase 4: prove that the page deserves to exist
- Assign one real reader task. A page should teach, compare, calculate, diagnose, document, or help complete a decision.
- Answer the task completely. Include prerequisites, steps, decision points, failure modes, and next actions where they matter.
- Separate evidence from opinion. Cite primary sources for standards, legal frameworks, product behavior, and measurable thresholds.
- Add original utility. A worksheet, test sequence, example, decision table, or worked scenario is more useful than a recap of other pages.
- Remove filler. Delete paragraphs that merely restate the heading, praise the topic, or repeat a target phrase.
Phase 5: create paths to the next useful answer
- Link every guide to its topic hub. The hub should explain the scope and expose the related sequence.
- Add contextual internal links. Link to a prerequisite, deeper explanation, or next step at the point where the reader needs it.
- Use descriptive anchors. Prefer "set up Search Console" over "click here," while avoiding repetitive exact-match phrasing.
- Keep navigation consistent. A reader should always be able to reach topics, guides, tools, policies, and contact information.
- Check every link. Validate internal destinations and make sure external citations still support the surrounding statement.
Phase 6: validate experience and measurement
- Test mobile layouts and keyboard access. Menus, forms, focus states, tables, and controls must work without precise pointer input.
- Measure performance. Follow the website speed optimization checklist with both field and lab data.
- Add only supported structured data. Schema must match visible content. Validate the JSON syntax and required properties.
- Generate an XML sitemap. Include canonical, indexable URLs and advertise the sitemap in robots.txt.
- Set up monitoring. Use the beginner Search Console workflow to submit the sitemap, inspect representative URLs, and watch indexing and performance after launch.
A two-pass launch test
| Pass | Question | Evidence to retain |
|---|---|---|
| Technical | Can a crawler fetch, render, canonicalize, and discover every important page? | Status report, crawl output, sitemap count, schema validation |
| Editorial | Would each page still be useful if it never ranked or linked to a commercial site? | Purpose statement, sources, reviewer notes, internal-link rationale |
If the business serves a defined geographic area, continue with the local SEO checklist for service businesses. Local visibility depends on accurate business information and service relevance, not a collection of near-duplicate city pages.