How to manage 50 WordPress sites
The work splits into two piles, and almost everybody solves one of them and then wonders why they are still drowning.
First, what the work actually is
At five sites you remember everything. At fifty you do not, and the failure is not dramatic: no single site is neglected enough to notice, and all of them drift.
Write down a month and it falls into two piles that need completely different answers.
| Pile one: keeping them alive | Pile two: changing them |
|---|---|
| Core, plugin and theme updates | A new landing page |
| Backups, and a restore when one breaks | A price or phone number on eleven sites |
| Uptime and SSL expiry | A section that broke on mobile |
| Malware scans | Three hundred product descriptions |
| The monthly client report | A page that takes six seconds |
Pile one is repetitive and identical across sites. Pile two is bespoke and per site. A tool that solves one is close to useless against the other, which is why agencies buy a maintenance dashboard, feel better for a month, and are still working weekends.
Pile one: buy a dashboard, and stop thinking about it
This is solved. Do not build it, do not script it, and do not do it by hand at fifty sites.
- ManageWP is hosted, free at the base, with the useful parts as per-site addons. Easiest to start.
- MainWP is self-hosted on a WordPress site you own, with no per-site fee. The right answer if you do not want a third party holding credentials.
- InfiniteWP is the other self-hosted option, a free core with paid addons.
All three do updates, backups, uptime and client reports across every site from one screen. Pick on hosted versus self-hosted and on how the pricing behaves at your site count, not on feature lists, because the lists are nearly the same.
We compare each of them against this product, and the conclusion on all three is to keep using them.
Standardise, or the second pile never shrinks
This is the part nobody wants to hear and the one with the largest return.
- One page builder. Fifty sites across Elementor, WPBakery, Divi and raw blocks is four products to stay fluent in, and every hire has to learn all four. New builds get one. Old sites keep what they have until they are rebuilt.
- One hosting arrangement, or two. Eleven hosts means eleven control panels, eleven support queues and eleven ways PHP is configured.
- A plugin baseline. A fixed list every site gets, and a short list nobody may add without asking. Most emergencies at scale come from a plugin somebody installed once.
- A written record per site. Which builder, which host, what is unusual, what broke last time, who the client is. Anywhere, as long as it is the same anywhere for every site. Half of what takes an hour at fifty sites is remembering.
This is slow and it compounds. An agency that did it three years ago runs fifty sites with two people; one that did not runs thirty with four.
Pile two: the part that has no dashboard
Once updates and backups are automatic, the week is made of the other pile — and none of it is repetitive in a way a bulk tool can address. Every one of those jobs needs somebody to open a site and understand what is on the page.
There are only three real answers, and it is worth being honest that the first two are what most agencies choose:
- Hire. Works, and it is the expensive answer. It also scales linearly, which is the thing you were trying to escape.
- Say no more often. Genuinely underrated. Not every client request is worth doing, and a retainer that excludes ad-hoc build work is a fair retainer.
- Automate the editing itself, which until recently was not possible, because editing a page builder means understanding its structure rather than its HTML.
That third one is what this product is. An AI agent you already pay for gets a locked-down set of tools for the sites you have linked, and the work happens on the live site: build the page, change the same section across twenty sites, rewrite the catalogue, find why the page is slow.
It does not do pile one. No scheduled backups, no uptime monitoring, no client reports. Keep the dashboard. The honest description is that it removes the reason you were about to hire.
Every screen of it is here, and what it refuses to do matters more than what it does when the sites belong to clients.
What changes at 50, specifically
| Thing | At 5 sites | At 50 sites |
|---|---|---|
| Knowing a site is down | The client rings | Monitoring, or you find out on Monday |
| Updates | Half an hour a week | A part-time job, or a dashboard |
| Credentials | In your head | A password manager, per site, revocable |
| Knowing what a site is | You remember | Written down, or it costs an hour |
| A bad change | You fix it | You need a backup from before it |
| One change everywhere | An afternoon | A week, or automation |
The credential problem, which gets worse quietly
At fifty sites you are holding fifty admin logins, often in a spreadsheet, often shared with whoever is helping this month. Nobody decided that; it accumulated.
The fix is cheap and nobody does it: stop using admin logins for tools. WordPress has application passwords for exactly this — one per tool, per site, revocable individually from the client's own profile screen, without changing anybody's login. Rotating one does not lock a human out of anything.
In order, if you are starting today
- Put every site on a maintenance dashboard. One afternoon.
- Write down what each site is. Boring, and it pays back within a month.
- Move tools off admin logins and onto application passwords.
- Standardise new builds on one page builder and one host.
- Only then look at automating the build work, because automating a mess produces a faster mess.