What happens after a website launches

Launch is treated as the finish line by everyone involved. The invoice closes, the team moves on, and the site goes quiet.

Then a year passes. The site is out of date, nobody remembers who has the hosting login, and a small change needs a quote. None of that is bad luck. All of it was decided in the week before launch, when nobody wanted to slow down a project that was finally nearly finished.

Agree these before going live

Who owns what. The code, in a repository with your name on it. The domain, registered to you. The hosting, billed to your card. Analytics, search console, and any third-party account the site depends on.

If any of these sits with the agency, you cannot leave without their cooperation. That is a commercial position, and it is usually not malicious — it is just how it ends up when nobody asks. Ask before launch, when you still have leverage.

What a small change costs. Get a real example, in writing. A new service page. A changed price. If each is a quoted project, you will stop asking, and the site will go stale.

What happens when something breaks. Not if. A form stops sending, a certificate expires, an update breaks a layout. Who do you call, how fast do they answer, and is it included or billed?

The handover matters more than the launch. A site you cannot change is a site that gets worse every week you own it.

The first month

Two things worth doing that almost nobody does.

Check the forms actually arrive. Submit each one and confirm the email lands — not in spam, in the inbox someone reads. Forms fail silently more often than any other part of a website, and the failure mode is invisible: no error, no bounce, just an enquiry that never existed as far as you know.

Check what search engines see. Confirm the site is indexed, the sitemap is submitted, and no page is accidentally blocked. A launch that ships with the staging robots rules still in place is a real and common way to be invisible for months.

The first year

Publish something, occasionally. Not to a schedule. When a customer asks a question worth answering properly, answer it as a page. Four of those a year is a real content library in three years.

Update what exists. A page that ranked and has not been touched will slip. Updating an existing page usually outperforms writing a new one, and nobody does it because it feels like less work.

Look at what people do. Once a quarter, look at which pages get visited and which produce enquiries. Usually one page does most of the work and nobody knew which.

Keep it patched. Whatever it runs on ships security updates. Unpatched sites get compromised, and a compromised site is removed from search results until it is cleaned — which takes weeks.

The retainer question

Most agencies offer one. Whether it is worth it depends entirely on what is in it, and the honest split is:

Worth paying for — security updates, backups, uptime monitoring, and a named person who answers when something breaks.

Usually not worth paying for — a monthly report nobody reads, and "content updates" capped at an hour you never use.

Ask what happened for the last client on the same retainer, last month. If the answer is vague, the retainer is insurance rather than service. That can still be fine, as long as you know which you are buying.

The one thing to insist on

Before the final payment: a document listing every account, every login, and where everything lives.

It takes an hour to produce and it is the difference between owning a website and renting one from people whose phone number you may lose. Ask for it at the start, not at the end.