Every site that stays up has somebody doing this work. Usually it is whoever has the password, doing it late, in a hurry, after something has already broken.
This is the least glamorous thing on the site and I am not going to dress it up. No state machine, no agent, no findings ledger. It is updates, backups, patches, a hosting account that renews before it lapses, email that lands in inboxes instead of spam folders, and the small content change you asked for on Tuesday being live on Tuesday. Invisible when it is done right. Very visible the week nobody is doing it.
What I keep running
- WordPress site management. Core, plugins, themes, and the version drift that turns into a broken layout six months after launch.
- Hosting, set up and kept running. Set up properly once, then watched, renewed, and moved when it needs moving.
- Email. Domains, DNS, deliverability. The part everyone assumes is finished until a client says they never received the invoice.
- Maintenance. Updates, backups you could actually restore from, security patches applied before a scanner finds them for you.
- Content and small changes on request. A new page, a new team member, a changed price table, a swapped image. Small asks that stop being small once they queue up for a quarter.
What is running today
Two clients are on this right now. One is on the full stack: site, email, hosting, and the ongoing services around them. The other is on maintenance and updates only, the steady hand that keeps a live site healthy without me touching anything else.
Both are real, ongoing retainers. Neither is named here, because client names stay generic on this site unless the client tells me otherwise. Pick the level that matches what your site actually needs rather than the bigger one.
This is not the AI ops retainer
Worth being plain about, because the two get conflated. Fractional AI Ops Lead is for a production AI system that drifts: a monthly audit of what the thing actually produced, one new capability a quarter, a warning when a model or your data moves. This page is for a website that has to stay up and keep working. Different work, different worry, different number.
If your site also runs an AI system, that is two retainers, and I would rather say so than fold one into the other and bill you for a monthly audit of nothing.
What happens when I am not there
A retainer here starts with a written list of the accounts under your site. Registrar, DNS, host and mailbox each get recorded with the name they are held in. Before you sign, not after. Where one sits in my name because that is how it arrived, the list says so, and the route back to you is written beside it.
The response window is a term in the agreement, not a mood. You get the number in writing before you sign, and it is sized to a bad week rather than a good one.
An AI Operations Executive joined full time in August 2026. Not cover for your site yet. Outward contact still comes through me, and the wiring that would let somebody else pick up a queue is half built: notifications that did not land, an access flag set wrong, work sitting with nothing telling anyone it was their turn. I will say that rather than sell you a rota.
Your site is not stored in my head. Tasks, approvals, who is allowed to touch what, and an audit row on every change all run through the operations platform I built for my own business. Every judgement call behind it gets written down as it is made, and that record is not something you receive. It is what makes a second operator possible here at all.
What this is not
Not a rebuild. If the site itself is the problem rather than the upkeep, start with Operations Infrastructure, get the thing right, then put it on a retainer.
Not an always-on SLA either. The window above is the one I hold, and it is written into the agreement rather than implied by it.
Not for everyone. One site I built ended cleanly at handoff, code and hosting signed over, the client fully independent, and that was the right ending for that project. If you already have someone who does this well, keep them.
Pairs with Operations Infrastructure when the site needs building or modernizing first, and with System Audit when you want to know what is wrong with what you already have before committing to keeping it running.