Website Launch Checklist
Pre-launch checks that actually bite: a staging robots.txt, a stray noindex, untested forms, missing analytics, expiring certificates and who owns the domain.
The interactive tool loads here when JavaScript runs.
About this toolWhat is the Website Launch Checklist?
Launch problems are rarely about design, because design is what everyone looked at. They are a robots.txt left over from staging, a noindex tag on a template, a contact form nobody submitted, a certificate that expires in six weeks, or analytics that was installed but never fired. Each is invisible on the day and expensive afterwards.
This list is grouped by the question being answered: does it work, can it be found, is it safe and legal, and will you know when something breaks. Work down it on the live site rather than on staging, because several of these items only have meaningful answers in production.
- Grouped by question: does it work, can it be found, is it safe, will you know when it breaks
- Covers the single-line failures that hide an entire site from search
- Includes the legal items most launches forget: an accurate privacy policy and consent
- Links the Security Headers Analyzer for the headers generated sites rarely set
- Editable and exportable as PDF, CSV, Markdown or a Google Form
Updated
How to use the Website Launch Checklist
- Run the list against the live site, not staging. Several items have different answers in production.
- Start with the findability section if you only have an hour: a blocked robots.txt is the one failure that hides everything else.
- Submit each form yourself and confirm the message arrives where someone will read it.
- Delete items that do not apply, such as cookie consent for a site that stores nothing, and add your own.
- Export the remaining work as Markdown to paste into an issue tracker, or as a PDF for a client sign-off.
Worked examples
Launching tomorrow
Input
Essentials only
Output
Forms, robots.txt, noindex, HTTPS, analytics, ownership
The items that are hard to recover from once traffic arrives.
Site is live but invisible in search
Input
Findability section
Output
robots.txt, noindex, sitemap, canonical, titles
Almost every "we are not in Google" case is one of these five.
Handing over to a client
Input
Full list, export PDF
Output
A signed-off record including who owns what
The ownership item prevents the most common post-handover emergency.
Background and insight
The findability items deserve the most attention because their failure mode is silence. A Disallow line inherited from staging removes an entire site from search, and nothing about the site looks wrong; you simply never appear. The same is true of a noindex meta tag, which hides pages individually and so survives casual checking. Both are single lines, and both are among the most common launch incidents there are.
Duplicate hostnames are the quieter version of the same problem. If a site answers on both the apex domain and the www subdomain without a canonical pointing at one, search engines treat them as competing copies and your signals are split between them. The fix is a redirect or a canonical, and the time to do it is before anyone links to the wrong one.
The legal items have become sharper. A privacy policy that names services you do not use is worse than having none, because it is a specific written claim that is false. And anything stored on a visitor's device beyond what the service strictly requires generally needs consent first, which in practice means analytics, since analytics is not strictly necessary for delivering a page.
Finally, ownership. The worst moment to discover that your domain sits in a former contractor's personal account is the day it expires. Writing down who holds the domain, the DNS and the hosting account takes a minute and prevents a category of outage that no amount of monitoring will catch.
Common use cases
Freelancers and agencies
A repeatable pre-launch pass and a document to hand over with the site.
Small business owners
The checks to request from whoever built the site, in plain terms.
Relaunches and redesigns
Where stale staging settings and lost redirects do the most damage.
Apps built with AI builders
Generated sites rarely set security headers or distinct page titles.
Tips and limitations
- Check robots.txt first. One inherited Disallow line removes the whole site from search and nothing looks wrong.
- Submit every form yourself. A silently failing contact form is indistinguishable from having no enquiries.
- Pick one of www or non-www and canonicalise the other, or your ranking signals are split in two.
- Trigger a real pageview and confirm it arrives, rather than trusting the analytics setup screen.
- Write down who owns the domain, DNS and hosting. Discovering it at expiry is the classic outage.
Sources and further reading
- Search Essentials, Google Search Central. What a site must do to be indexable
- Introduction to robots.txt, Google Search Central. The directive that can hide an entire site
- Web Content Accessibility Guidelines (WCAG) 2.2, W3C. The accessibility baseline to check before launch
Frequently asked questions
What should I check before launching a website?
In order: that forms and the main journey work on a real phone, that robots.txt and meta tags are not blocking search engines, that HTTPS and security headers are in place, and that analytics and uptime monitoring are actually reporting. Then record who owns the domain and hosting.
Why is my new website not appearing in Google?
Most often a Disallow rule in robots.txt inherited from staging, or a noindex meta tag left on a template. Both are single lines, both hide pages completely, and neither makes the site look broken. Check those before considering anything about rankings.
Do I need both www and non-www versions?
Your site may answer on both, but only one should be canonical. Serving both without a canonical or redirect means search engines treat them as separate competing copies and your signals are split between them.
Is a generated privacy policy good enough?
Only if you edit it to match reality. Templates routinely name analytics or payment services you do not use and omit ones you do, which makes it a specific written claim that is false. List every third party that receives visitor data and match the policy to that list.
What should I check after launch rather than before?
That analytics recorded real visits, that the uptime monitor fires when you deliberately take a page down, and that the first search-engine crawl indexed what you expected. Search Console will tell you which pages were refused and why.
Is there an in-depth guide to the Website Launch Checklist?
Yes. “Why Checklists Work, and How to Use One Properly” covers the background, how it works, common pitfalls and worked examples. Read the full guide.
Related Tools You May Like
Free, in-browser tools that pair well with this one.