A good launch is boring, and that's the point. The drama — dead links, missing meta, a contact form that emails nobody — is almost always avoidable. This is the list we run before a site goes live. Three of the items are on it because they went wrong for us first.
Content and polish
- Proofread everything. Real content, no "lorem ipsum", no placeholder phone numbers. Read it aloud once.
- Click every link and button — especially the calls to action. Nothing kills trust like a dead "Contact us".
- Test the contact form end to end. Submit it for real and confirm the email arrives in the right inbox, not spam.
SEO and sharing
- A unique title and meta description on every page. Not "Home | Home".
- Open Graph tags and a share image, so a link pasted into Facebook or LinkedIn looks deliberate rather than broken.
- A sitemap, and permission to be indexed. A "noindex" left over from the staging site will quietly keep a finished site out of Google.
Performance and mobile
- Compress the images. The biggest, cheapest speed win — no 4 MB hero photos. Why it's first on the list.
- Test on a real phone. Tap the menu, fill in the form, check the buttons are thumb-sized.
- Check more than one browser. A pass in Safari and Chrome catches most surprises.
Security and launch day
- HTTPS everywhere, a valid certificate, and old http:// addresses redirecting to https://.
- Backups and monitoring switched on before launch, not after the first incident.
- Redirects from the old site's addresses if you're replacing one, so the search rankings it earned come across with you.
The three we learned the hard way
Check what the server is publishing, not just what the site shows
In August 2026 we found that a deploy setting on our own site was publishing the entire code repository, not just the website. Anyone who guessed a path could download the server code and the layout of the database. No passwords or keys were in it — but it didn't matter whether anyone had found it; it shouldn't have been there. The site looked perfect the whole time, which is exactly why a visual check would never have caught it. Our build now publishes a fixed list of files and refuses to deploy if anything outside that list turns up.
Test every address on the live host
Our free audit page once returned "page not found" on the live site. The rule that made its short address work sat in a config file that the host didn't apply on a normal deploy. Load the real URLs, on the real host, after the real deploy — not on your laptop.
Check your contact details with scripts blocked
In September, a security service in front of our own domain automatically rewrote an email link to hide the address from spam bots. With scripts running, visitors saw the address normally. With a privacy extension blocking the script, they saw the words "[email protected]" and a link that went nowhere. Open the live site once with JavaScript switched off and make sure a customer can still find a way to reach you.
The best launches are the ones nobody remembers, because nothing went wrong. That calm gets built the day before, with a list.
Running all of this on every site is repetitive, and it's the sort of thing a well-run platform should handle by default — HTTPS, backups, monitoring, sitemaps and sensible SEO come switched on. If you'd rather have it built for you, this is the list we start from on every website we design in Ballarat. If you'd rather build it yourself, the platform below has it switched on already.
Want launch-ready sites with the boring bits handled?
Explore Websites →