Technical SEO checklist for a new website: 15 pre-launch checks

A technical SEO checklist for a new website is a set of pre-launch checks that confirm search engines can crawl, index and understand every page. It also covers redirects from old URLs, mobile speed and analytics that runs only after cookie consent. Technical SEO setup is included in our brand and eCommerce website packages. Use the 15 checks below as a pre-launch website checklist, then run them again on the live domain as your website go-live checklist.

So what is included in technical SEO? The checks fall into five groups: indexing (1–4), URLs and redirects (5–7), on-page and structured data (8–10), speed and mobile (11–12), and analytics, consent and AI search (13–15). Each gives the check, why it matters, how to verify it and who does it. The pages and legal details a small business website needs are covered in a separate guide.

1. 01Robots.txt allows crawling – a leftover “Disallow: /” from staging hides the whole site. Verify: open /robots.txt on the live domain. Check it yourself.
2. 02No leftover noindex – a staging noindex tag keeps pages out of Google. Verify: URL Inspection in Search Console. Ask your developer.
3. 03Self-referencing canonical – each page names its own preferred address, so duplicates do not compete. Verify: rel="canonical" in the page source. Ask your developer.
4. 04XML sitemap – only live, indexable pages, submitted in Google Search Console and Bing Webmaster Tools; one sitemap file holds up to 50,000 URLs or 50 MB (Google Search Central). Verify: /sitemap.xml and the Sitemaps report. Ask your developer.
5. 05One 301 per old URL – each old address goes in one step to its closest new page. Verify: curl -I or a redirect checker. Ask your developer.
6. 06One URL form – http, www and trailing-slash variants land on one address. Verify: type each variant into a browser. Check it yourself.
7. 07Real 404s and crawlable links – missing pages return a 404 status; menus and in-text links are plain HTML links. Verify: a made-up URL and the page source. Ask your developer.
8. 08Unique titles and descriptions – a title of up to about 60 characters and a meta description of up to about 155 on every page. Verify: the browser tab and page source. Check it yourself.
9. 09One H1 and alt text – one clear main heading per page, headings in order, images described in words. Verify: the page source. Ask your developer.
10. 10Structured data and sharing image – schema repeats only visible facts; the 1200 × 630 Open Graph image loads. Verify: Google's Rich Results Test, and open the og:image URL to confirm it returns 200. Ask your developer.
11. 11Core Web Vitals on mobile – slow pages lose visitors, and Google's ranking systems use these metrics. Verify: PageSpeed Insights on mobile, median of three runs. Ask your developer.
12. 12Mobile layout and images – no sideways scrolling, readable text, easy-to-tap buttons, WebP or AVIF images. Verify: browse the site on your own phone. Check it yourself.
13. 13Analytics only after consent – UK rules require consent for most analytics cookies. Verify: no analytics cookies in a private window before you click Accept. Check it yourself.
14. 14Conversions and account ownership – forms and phone and email clicks are tracked; the business owns Search Console and Google Analytics. Verify: the GA4 real-time report and user lists. Check it yourself.
15. 15Readable by search and AI crawlers – text, FAQs and links in the initial HTML; no robots.txt or firewall rule blocks search bots. Verify: view-source or curl. Ask your developer.

To get a new website on Google, make sure nothing blocks crawling and submit an XML sitemap in Google Search Console. Then request indexing for your most important pages with the URL Inspection tool. Google then has to crawl the pages before they can be indexed and become visible in search results.

SEO for a new website often goes wrong on launch day, and most of those problems come from staging: a robots.txt “Disallow: /” rule or a noindex tag from the development copy is carried over to the live site by mistake. Protect staging with a password instead, and check the live robots.txt and page source as soon as the domain switches over.

Give every page one canonical address – for example https, without www and with a trailing slash – and use that exact form in canonical tags, the sitemap and internal links. Keep the sitemap's lastmod dates real: they should change only when the content does.

Once the domain is live, follow these steps:

1. 01Check that the live robots.txt does not block crawling and that no page carries a leftover noindex tag.
2. 02Verify the domain in Google Search Console and Bing Webmaster Tools.
3. 03Submit your XML sitemap in both tools.
4. 04Request indexing for your most important pages with the URL Inspection tool.
5. 05Open the Page indexing report if a key page is still missing after several weeks.

Google says crawling can take anywhere from a few days to a few weeks (Google Search Central, checked 23 October 2026).

URLs are ready for go-live when every page has one readable, lowercase address and every old URL redirects to its closest new page in a single 301 step. Internal links should point straight at the final addresses. Redirect chains and duplicate versions of the same page split signals between addresses and waste crawling.

Map every old URL – from the old sitemap, Search Console and analytics – to its closest new page, not the homepage, and test a sample with curl -I: one permanent redirect (301 or 308), then a 200. Unknown addresses should return a real 404 status, not a 200 “not found” page, which Google treats as a soft 404.

Menus, footers and in-text links should be ordinary <a href> links in the HTML the server sends, with link text that describes the destination – “Read more” tells search engines nothing.

Titles, headings and structured data match each page when every page has its own hand-written title and meta description and one clear H1. Structured data should repeat only facts visible on the page. On new websites these are often left on template defaults.

• Title: unique, about 60 characters or fewer, main topic first and brand name at the end.
• Meta description: about 155 characters or fewer, written by hand to say what the page offers and why to click – not its first lines cut off.
• Headings: one H1 written as continuous text, then H2 and H3 in order, with no headings used only for styling.
• Images: alt text describes what is in the picture; purely decorative images get an empty alt.
• Structured data: Organization, BreadcrumbList and, for articles, BlogPosting with real published and updated dates; FAQPage only when the same questions and answers are visible on the page.
• Sharing image: a 1200 × 630 Open Graph image that returns a 200 status, so shared links show a proper preview.
• Meta keywords: leave the tag out – Google does not use it for ranking.
• Hreflang on bilingual and multilingual websites: each language version links to the others, plus an x-default version. The bilingual WordPress website we built for a Klaipėda athletics school in Lithuania carries hreflang for Lithuanian, English and x-default (checked 23 October 2026).

A new website is fast enough on mobile when its key pages pass Google's “good” Core Web Vitals thresholds, starting with Largest Contentful Paint within 2.5 seconds. The other two are Interaction to Next Paint of 200 milliseconds or less and Cumulative Layout Shift of 0.1 or less, measured at the 75th percentile of page loads (web.dev, checked 23 October 2026).

Before launch there is no real-user data, so test the homepage, a main service or category page and a typical article or product page in PageSpeed Insights on mobile, three runs each, and take the median. After launch, field data appears in Search Console once enough real visits are recorded.

Typical fixes: WebP or AVIF images with width and height set, the main image loaded early, no loading screen or hero hidden with opacity: 0 until JavaScript runs, video loaded only when the visitor presses play, and self-hosted or preloaded fonts. Still choosing a platform? See whether Next.js is faster than WordPress.

Analytics and cookie consent are set up correctly when no analytics or marketing tag runs before the visitor makes a choice, and rejecting is as easy as accepting. Under PECR, non-essential cookies – including most analytics cookies, such as Google Analytics – need the visitor's consent before they are set. The Data (Use and Access) Act 2025 adds limited exemptions; check the ICO's current guidance before relying on them (checked 23 October 2026).

Test it in a private window: with the developer tools open, no analytics cookies should appear before you click anything, and nothing should change after Reject. Accept and Reject should be equally prominent, with no extra steps to refuse.

Put conversion tracking on your new website launch checklist: track form submissions and phone and email clicks from day one, and send a real test enquiry through every form. The business, not the developer, should own the Search Console property and Google Analytics account, with the developer added as a user.

Search and AI crawlers can read a new website when its text, FAQs and links are in the HTML the server sends and nothing blocks them. Check view-source or run curl on a key page: if the main text is missing there, many crawlers will not see it either.

Make sure robots.txt and firewall rules allow Googlebot, Bingbot, OAI-SearchBot, PerplexityBot and Claude-SearchBot. Put FAQs in plain HTML, such as <details> elements, with matching FAQPage data, and keep prices, addresses and opening hours identical everywhere. An llms.txt file is cheap, but its benefit is unproven – treat it as optional.

After launch, re-check the live website on day 1, day 7 and day 30; there is no set time for a new website to rank on Google. Nobody can guarantee positions, so do not expect a fixed date for rankings. Specific, less competitive searches usually start to show results before broad, competitive ones.

Technical SEO setup is included in our brand and eCommerce website packages. After launch, Complete Digital Care keeps the website monitored 24/7, with support Monday–Friday, 09:00–18:00.

Planning a new website or a rebuild? Get a quote – we reply within 24 business hours.

• Day 1: run the whole website launch SEO checklist – all 15 checks – on the live domain, test old URLs against the redirect map, submit the sitemap and send a test enquiry through every form.
• Day 7: in the Page indexing report, look for pages excluded by noindex, 404s from old links and pages crawled but not indexed, and fix the cause.
• Day 30: review which pages and searches receive impressions in the Performance report, compare old and new URLs, and check Core Web Vitals once field data appears.