Advice

The go-live checklist we run on every website

Every site we launch goes through the same go-live checklist, Statamic or WordPress. Favicon in place, passwords hardened, transactional email routed through Postmark, images on a CDN, SEO set up at launch, redirects mapped for the old URLs, analytics wired up properly and the whole thing deploying from Git. Here is what the checklist actually covers, and what you get with managed hosting that a budget shared plan quietly leaves out.

Brad Goddard Brad Goddard · Founder, We Are Jungle · 20 November 2026
wearejunglebrandshoot-iggy&lime-205.jpg

There is a moment on every build where the site stops being a project and starts being live. The DNS flips, somebody somewhere refreshes the page and your new website is suddenly a real thing people can type into Google and find. The job before that moment is the bit nobody sees, and it is the bit that quietly decides how the next twelve months go.

We run the same go-live checklist on every site we launch, whether it is Statamic, WordPress, Shopify or a custom build. Some of it is finishing touches, some of it is plumbing that will matter in six months and some of it is the quiet stuff that stops a cheap shared host costing you money later. Here is what it actually covers.

The tiny finishing touches

The small bits are what give a site away. A missing favicon, so your tab in the browser shows a grey outline where everyone else has a proper logo. An Open Graph image that has not been set, so when somebody shares the page on LinkedIn it comes up blank with a URL and a question mark. A page title that still says "Home" because nobody got around to writing one. None of these break the site. All of them look amateurish, and all of them are five minute jobs that somehow never get done if there is not a checklist.

We set the favicon in the right sizes for modern browsers and iOS home screens, write the page titles and meta descriptions, pick a sensible Open Graph image for sharing and then sanity check how the site looks when somebody drops the URL into WhatsApp or Slack. If you have never actually tested that on your current site, go and try it. You will usually learn something.

Passwords, done properly

Every account that touches your site gets a strong, randomly generated password. Not the one you use for everything. Not the clever one with a 1 on the end. Hosting, CMS, email sending, analytics and anything else with a login all get unique credentials stored in our password manager, with access given to whoever on your team actually needs it. If somebody leaves, we rotate what needs rotating.

The reason to bother is that reused credentials are still the easiest way into a small business website. Verizon's Data Breach Investigations Report has flagged stolen credentials as the top way into a breach for years running. Small sites are almost never targeted on purpose. They are caught in automated sweeps, and a reused password is the open door those sweeps are looking for. Twenty minutes of setup at launch saves a weekend of grief later.

Transactional email that actually arrives

This is one of the biggest differences between a proper launch and one that is cobbled together. Every contact form on your site, every password reset, every order confirmation, every booking notification. All of it needs to arrive in an inbox, not a spam folder. The cheap way to send those emails is straight from the web server the site is on, and it is also the way almost nothing arrives at anyone using Gmail, Outlook or iCloud.

We route all transactional mail through Postmark, a dedicated sending service built specifically for the one-off emails sites fire off. Their published delivery rates to the inbox sit around 99 percent, and the few that do not land tell you why, so you can fix the cause rather than guess. On top of that we set up SPF, DKIM and DMARC records on the domain itself so Gmail and Outlook actually believe the email is coming from you. Those three acronyms are the reason your old form submissions used to disappear.

Images served the way they should be

For Statamic builds we push images to object storage with a CDN sat in front of them, rather than letting them live in the site itself. There are two reasons. First, images are usually the single heaviest thing on a page. HTTP Archive's 2024 Web Almanac puts them at roughly half of total page weight on a typical site. Serve them from a cheap shared host in one country and somebody browsing from another is quietly waiting.

Second, separating images from the site makes the site itself much smaller and much easier to deploy, back up and move. If we need to spin up a staging copy of the site, the staging copy does not have to drag the entire image library along with it. For WordPress we handle the equivalent at the hosting layer rather than bolting another plugin on top of the pile.

SEO set up at launch, not six months later

The SEO plugin goes on while the site is being built, not six months after launch when somebody wonders why the site is not ranking. Meta titles and descriptions are written per page, schema is in place for things like services, team members and articles, canonical tags are set correctly, the sitemap is submitted to Google Search Console and we check there are no accidental "noindex" tags left on from staging. That last one has caught more sites than anyone would like to admit.

Statamic and WordPress handle this a little differently under the bonnet, but the result is the same. The site is already saying the right things to Google on launch day. There is no magic SEO setting that outranks poor content, but there is a handful of settings that will quietly stop a good site ranking if they are wrong. We get those right before the site goes live. If you are curious what that looks like for a site that is already live, our free website audit walks you through it in about fifteen minutes.

A 301 redirect map for the old URLs

If we are building you a new site to replace an old one, we write a 301 redirect map before the switch. Every URL that is changing gets mapped to its new home. The old /about-us/our-team becomes the new /about/team, the blog post that used to be /news/42-widget-launch becomes /blog/widget-launch and anything genuinely retired points at the closest sensible page.

This matters because Google's own guidance on moving a site is quite plain about it. Migrate without proper redirects and expect to lose rankings, often for months. The authority a page has quietly built up over the years is tied to its URL. Change the URL and lose the redirect and you throw that away. We would rather spend half a day writing the map than watch six months of rankings disappear on launch day.

Analytics that will still be there in a year

Google Analytics 4 goes on through Google Tag Manager, not pasted directly into the site code. The reason is practical. Tag Manager lets us add, change or swap out analytics, conversion tracking, paid ad pixels and anything else that fires on the page, without going back into the site every time. Six months after launch when the marketing team wants to measure a new campaign, that saves a developer invoice.

The other reason to do it on launch day rather than later is that your Analytics view starts counting from day one. Set up six months in, you have six months of invisible traffic nobody can get back.

Deploying from Git, not dragging files over FTP

Every site we build lives in a Git repository on GitHub, and deployments happen automatically when we push changes. There is no FTP, no "did I upload the right file", no mystery changes nobody can quite explain. If something needs rolling back we roll back to a specific commit, not a copy of a folder somebody zipped up last Tuesday. Every change on the live site has a name against it and a timestamp.

This is the bit of a modern build that cheap shared hosting actively gets in the way of. A budget plan usually gives you an FTP login and a WordPress dashboard and calls that deployment. A professional build needs more than that, and the hosting environment needs to support it. Ours does.

SSL, backups and server-level caching

SSL certificates are issued, renewed and monitored, so a certificate never quietly expires overnight and takes the site offline for the morning. Backups run automatically, retained for long enough to be useful and, importantly, actually tested. There is no point in a backup you have never restored. Caching sits at the server and the CDN layer, not bolted on with a plugin that fights the theme. All of it is checked before launch and quietly watched afterwards. The PHP2Shell week was a decent illustration of what that quiet watching actually does for you.

What hosting with us bundles that a budget shared host does not

If you are on a budget shared plan with somebody like GoDaddy or IONOS, there is nothing wrong with the company. They do what they advertise, which is sell cheap hosting at scale, usually with a slick signup flow and a dashboard full of upsells. What they do not do is any of the checklist above for you. The checklist is yours to work through, or more usually, yours to ignore until something breaks.

Hosting with us gets you the whole lot bundled with the people who built the site:

  • Someone who already knows how your site is put together, on the end of an email. Not a ticket queue, not a chatbot asking you to describe the issue in three to five words.

  • Managed UK infrastructure with edge caching and a CDN, tuned for the shape of the site rather than a one-size-fits-all shared plan.

  • SSL, DNS, backups and server-level security patching handled quietly in the background.

  • Transactional email routed through a dedicated service that actually delivers to inboxes rather than junk folders.

  • Analytics, redirects and SEO set up properly at launch rather than guessed at six months later.

  • Staging environments available when the site needs them, so plugin updates and bigger changes can be tested before they touch live.

Most of our clients bundle hosting into a website support plan. Same team, one invoice, one relationship. When something needs doing at the server level and the site level in the same afternoon, that stops being three different suppliers and three different tickets. If you want hosting on its own we also do that, usually for clients who have an in-house technical team already.

If you are thinking about moving off GoDaddy or IONOS

If you already have a site on a budget shared host and suspect it is holding you back, we move sites over without the drama. The usual pattern is: we spin up the new environment, copy the site across, point everything that needs pointing and schedule the DNS switch for an off-peak time. No downtime that visitors would notice. No weekend phone calls. The old host gets politely cancelled once we are confident the new one is behaving.

Even if the site itself is fine and you just want it on a hosting environment that has actually been thought about, we can migrate it, set up the parts of the checklist that were missing and bring it under a support plan so it stays looked after. What that ongoing support covers is a conversation in itself, but it starts with the checklist above.

The short version

A checklist is unglamorous. It is also the thing that quietly stops the irritating problems from happening at all. Favicon in place, strong passwords, email routed through a proper service, images on a CDN, SEO set up before launch, redirects mapped, analytics wired up, deployment running off Git and the hosting handled by people who built the site. None of it is clever on its own. All of it together is the difference between a site that quietly works and one that quietly does not.

If you are planning a new site or itching to move the current one somewhere better, come and have a chat. We would rather have the conversation before the launch than clean up after one.

Hosting

Managed Hosting Built Around Your Site

Explore Hosting
Share

Read Next

website-audit.jpg

Our 24 hours with PHP2Shell

On Wednesday the PHP team shipped an emergency fix for two critical flaws affecting roughly 70 percent of websites. Here is the honest story of our 24 hours, the false alarms, the frankly stupid amount of coffee and why none of our hosted clients had to lift a finger.

Read Blog