What a CMS actually costs

"We need an admin panel so we can change content ourselves" is one of the most common items in a brief. It sounds self-evidently useful. Let us do the arithmetic.

What makes up the cost

A CMS is not a checkbox in settings. For a corporate site with 15 page types it means:

• Modelling content schemas for every page type — 20–30 hours
• CMS setup, roles, permissions, localisation — 15–25 hours
• Wiring to the front-end, draft previews, rebuilds on publish — 25–40 hours
• Validation, so an editor cannot break the layout with an overlong heading — 10–15 hours
• Documentation and training — 8 hours

That is 80–120 hours: $1,400–2,300 and two to three weeks of schedule. Then CMS hosting, updates, and a new place where things can break that could not break before.

What to compare it against

The alternative is edits under a maintenance contract. Changing text on a page, adding a team member, publishing a post — 15 to 40 minutes of work. At our $55 hourly rate, roughly $25 per edit.

$2,300 ÷ $25 = 92 edits. The CMS starts saving money after the ninety-second change.

How many edits actually happen

We looked at 23 sites under our maintenance and counted content change requests over a year:

• Corporate sites with no blog: 4–11 edits a year
• Sites with a blog written by an agency: 30–60 a year
• Catalogues with changing prices: 200+ a year
• Property developer sites with changing availability: thousands — but that is a CRM integration, not a CMS

In the first group a CMS never pays for itself. In the second it takes two to three years. In the third it pays back in months and is genuinely required.

The less obvious argument against

A CMS used four times a year does not work. Someone logs in after six months, cannot remember where anything is, is afraid of breaking something, and emails us: "could you just change this bit". We have watched this happen repeatedly.

So you paid $2,300 and you are still emailing your contractor. Only now you are also paying to host a CMS.

What we propose instead

On sites with infrequent changes, content lives in the codebase and we make edits under maintenance within the business day. If edits become frequent a year later, a CMS can be added — the content is already modelled in schemas, so connecting one takes less work than building it up front would have.

This is not a "what suits us" position. Selling a CMS suits us perfectly well: it is 100 billable hours. But a project where the client paid for something unnecessary comes back a year later as dissatisfaction, not as a second contract.