Website Pre-Launch SEO Checklist: Make Sure Google Can Find and Read the Site

What can pre-launch SEO do for a new site, and what can it not?

Most pre-launch SEO is damage prevention. A new site rarely ranks for anything on day one, and nothing here promises that it will. What you can do is avoid the self-inflicted problems that keep a site invisible: a staging block left switched on, a title that is the same on every page, links Google cannot follow. Google’s own starter guide says that creating compelling and useful content will likely influence your site’s presence in search results more than any other suggestion in it, so treat this list as the plumbing around the real work.

Every statement below comes from Google Search Central pages read on 13 November 2026, linked in the sections that use them. Google updates its guidance, so confirm anything that matters on the page itself. If you want the longer strategy, the launch with SEO from day one playbook covers content and keywords, and this page stays with the checks.

Can Google reach the site at all?

Sites are often built on a hidden staging copy, and the blocks that kept it hidden travel with the code. The most common cause is simple: someone forgot to take them off. Check each of these on the live address, not on your development machine.

1. Search the page source for <meta name="robots" content="noindex"> and remove it from pages you want in search. Google’s noindex documentation says a page with the rule is dropped entirely from results, however many sites link to it.
2. Check the response headers for X-Robots-Tag: noindex, because a hosting setting can add it without touching your HTML. The same page says the header has the same effect as the tag.
3. Open /robots.txt and look for a Disallow: / left over from staging. Google’s robots.txt introduction says the file manages crawler traffic, and that it is not a mechanism for keeping a page out of Google.
4. Know which tool does which job. A page blocked by robots.txt can still appear in results without a description, and for a noindex rule to work the page must not be blocked, or the crawler never sees the rule.
5. Open the site in an incognito window. Google’s page on getting a site on Google says a login wall or a block can stop its crawler from seeing the content, and an incognito window shows what an outsider sees.
6. Check that the home page and each key page return a normal success response, not a redirect loop or an error.

# Status code and headers: look for 200 and for any x-robots-tag line
curl -sI https://example.com/

# Title, robots meta, canonical and description in the page source
curl -s https://example.com/ | grep -i -E "<title>|name=.robots.|rel=.canonical.|name=.description."

# robots.txt: it should not contain 'Disallow: /' on its own line
curl -s https://example.com/robots.txt

User-agent: *
Disallow:

Sitemap: https://example.com/sitemap.xml

Can Google understand each page?

• A unique, descriptive title on every page. Google’s title link guidance says to avoid vague text such as “Home”, to avoid keyword stuffing and to avoid repeating boilerplate across titles, since it makes pages indistinguishable in results.
• A meta description written for that page. Google’s snippet guidance says snippets come mainly from the page content, that Google sometimes uses the meta description when it describes the page better, and that identical descriptions across a site are not helpful.
• One clear main heading. Google says it looks at headings and prominent text when it builds a title link, so make the first h1 the real title.
• Text, not only images or video. Google’s getting-started page says to use text, not just images or video, on your site.
• Descriptive URLs. The starter guide suggests words in the URL that help users, rather than random identifiers.
• Real links. Google’s crawlable links guidance says it can generally follow only an a element with an href, and that anchor text should say what the target is. A button driven only by a script event may not be followed.

If your pages are built with JavaScript, use the URL Inspection tool in Search Console to see the HTML Google received. Google’s JavaScript basics page says you can set the title, description and canonical with JavaScript, but that the best place for the canonical is the HTML. The same check applies to the share card used elsewhere, covered in OG image specs.

Which redirects, errors and duplicates should you settle first?

Pick one version of the address and make every other version point to it. Decide between www and non-www, with or without a trailing slash, and the secure address, then redirect the rest. Google’s redirect guidance says 301 and 308 mean a page has moved permanently and are the right choice when the address changes for good, while a 302 is a weaker signal suited to a temporary move.

• Redirect once, not in a chain. Google’s status-code page says Googlebot generally follows up to 10 redirect hops, but a single hop is faster for people.
• Return a real 404 for a missing page, not a 200 with an error message on it; Google’s status-code documentation says Search Console shows a soft 404 error when a success response carries content that looks like an error or an empty page.
• Avoid two URLs with the same content. The starter guide says search engines choose one canonical URL, that duplicate content is not a spam violation, and that redirects or a rel="canonical" link help when you cannot avoid it.
• Fix staging leftovers. Search your pages for links to a staging or preview address and replace them with real ones.

Does the mobile version carry the same content?

Google’s mobile-first indexing guide says it uses the mobile version of a page for indexing and ranking, and recommends responsive design, which serves the same HTML on the same URL. It also says to use the same robots meta tags on mobile and desktop, because a different noindex or nofollow on the mobile page may stop it from being indexed, and not to lazy-load primary content only after an interaction such as a click or swipe, because Google will not perform it.

In practice, open the page on a phone and compare it with desktop. If a section of text, a link or an image appears only on one, it may not count. The landing page launch checklist covers the visitor’s view of the same screen.

How do you tell Google the site exists?

1. Add the site to Search Console and verify ownership, since it shows errors that stop Google understanding the site and lets you send important URLs.
2. Use the URL Inspection tool on the home page and a few key pages, to see what Google received, then request indexing for those. Google’s getting-started page says you can send your most important URLs to be crawled and potentially indexed.
3. Decide about a sitemap. Google’s sitemap overview says a small site, of around 500 pages or fewer and comprehensively linked, might not need one, but that a new site with few external links can benefit. A sitemap helps discovery and does not guarantee indexing.
4. Link every page from another page. Google’s starter guide says it mostly finds pages through links from pages it already knows.
5. Earn links from places where your readers actually are: a launch listing, a post in a relevant community, a write-up. Google’s page says not to pay for links, because that can break its spam policies.
6. Wait. The getting-started page says it can take a few weeks for Google to notice a new site, so check again before you conclude anything.

A launch listing is one way to earn a first link. On this site, the outbound link on a launch page to your own website is a normal followed link, while links inside text people write are marked nofollow and ugc. Launch your product free when the site is live, and see whether launch directories give backlinks for how to think about their value.

What is the order of work, and what do you leave out?

• Skip keyword stuffing in titles and text; Google’s guidance names it as something that makes a result look spammy.
• Skip buying links or directory packages sold on a promise of rankings; nobody can promise them.
• Skip structured data until the basics pass, unless you know exactly what it would enable.
• Mistake: copying a competitor’s text. The starter guide asks for content that is unique and written from what you know.
• Mistake: reading the first week as a verdict. A site:search returning nothing in week one is normal for a new site; Google says the operator does not necessarily return all indexed URLs anyway.

If you host the site yourself and want a free starting place, the guide on how to host a website for free lists options, and the ideas board shows what other builders are working on.