The site is finished and an invoice arrives for a monthly plan. It might be sixty a month or six hundred, the description is two lines long, and there is no obvious way to judge whether that is reasonable.
Two questions settle most of it. What is actually in the plan, and do you still own your website if you stop paying? The second one matters more and almost nobody asks it.
The short answer, by type of site
| Site | Typical monthly | What that should cover |
|---|---|---|
| Static brochure site on a reliable platform | $0 to $60 | Honestly, very little is needed. Hosting and a domain. Do not buy a plan you will not use. |
| Platform site with a handful of plugins | $80 to $300 | Updates and testing, backups, uptime checks, security patching, a small allowance of changes, and a human who answers. |
| Busy site, shop, or anything with integrations | $300 to $1,200 | All of the above plus monitoring of the paths that earn money, faster response, and someone who knows your setup rather than a queue. |
| Custom application | $1,000 to $5,000 and up | Dependency and platform upgrades, security work, incident response, and enough regular attention that it does not quietly rot. |
Illustrative ranges from the kind of work we quote, not a price list. Hosting is usually separate and usually small; the cost here is people, not servers.
What real website maintenance includes
- Updates, applied and tested. Not just applied. The value is that someone checks the site still works afterwards, because the update that breaks something is the reason to pay a person rather than to tick a box.
- Backups you have seen restored. A backup nobody has ever restored is a belief, not a backup. Ask when the last test restore was and how long it took.
- Security patching, with a stated approach to something urgent appearing on a Friday.
- Monitoring that reaches a human. Not a dashboard nobody opens. Somebody should know before your customers do.
- A stated response time, and clarity that response is not the same as repair. Responding in an hour and fixing in two days is a perfectly honest offer; discovering the difference during an outage is not.
- A small allowance of changes, with unused time either rolling over or being honestly described as use it or lose it.
- A named human. The single largest difference between a plan worth having and a plan that is a line item.
- A monthly note of what was done. Even three lines. Otherwise you are paying for something invisible and eventually you will resent it.
What a cheap plan leaves out
Rarely dishonest, usually just unstated, and the gaps are always the same ones.
- Testing after updates. Automated updates with nobody looking. You find out when a customer does.
- Anything urgent. Cover is business hours, which is reasonable, but you should know it before the Saturday your checkout stops working.
- Actual fixes. Some plans cover monitoring and updates but bill separately for repairing whatever they find, which can be a fair arrangement and needs to be said out loud.
- Content changes. Frequently excluded, so every text change becomes a small invoice.
- Restores. Backups included, restoring them chargeable. Ask.
- The platform itself. Major version upgrades are often outside a plan, and they are the expensive ones.
The question almost nobody asks
Here is where the real money is, and it has nothing to do with the monthly figure.
Some monthly arrangements are maintenance. Others are a subscription where the site is built at little or no upfront cost and the fee continues indefinitely, and if you stop paying, the site goes away. Both are legitimate businesses. They are wildly different purchases, and the two-line description on the invoice looks identical.
So ask, plainly, before signing anything: if I stop paying next month, what happens? If the answer is that you keep the site and lose the support, that is maintenance. If the answer is that the site comes down, you are renting, and the arithmetic changes completely. A plan at three hundred a month for eight years is a substantial sum for a website you will never own.
Then ask who holds the domain and the hosting account. Those should be in your name regardless of who maintains them, and who owns your website covers how to check and what to do if the answer is uncomfortable.
What actually goes wrong without one
Not scare stories, and the honest version is less dramatic than the sales version.
An unmaintained site does not usually explode. It decays. Plugins fall behind and stop being compatible with each other. The platform moves to a version your theme was never tested against. A certificate expires and browsers start warning people away. A form stops sending and nobody notices for weeks, which is the most expensive failure on this list because it is silent and it costs you enquiries you never knew you had. Eventually something needs fixing, and the person quoting for it has to first work out what has changed over three years, which costs more than the maintenance would have.
The security risk is real but usually less dramatic than described. Small business sites are rarely targeted deliberately; they are found by automated scanning for known unpatched weaknesses. Which is exactly what patching prevents, so the boring work is the work that matters. If it has already happened, my website was hacked is the recovery order.
The difference between keeping it alive and making it better
Worth separating, because plans mix the two and the mixture is where dissatisfaction comes from. Keeping a site alive is insurance: updates, backups, patching, monitoring. You are paying for nothing to happen, and when nothing happens it feels like paying for nothing.
Making it better is different work with a visible result: a page rewritten, a form simplified, a slow page fixed, something added because a customer asked twice. Most plans bundle a few hours of this and call them improvement hours, and two things go wrong with them.
- They go unused for months and then get spent badly. Hours that expire monthly quietly become a discount you never claimed. Ask whether they roll over, and if they do not, put a recurring reminder in your calendar to use them.
- They get treated as a free feature budget. A few hours a month is real capacity for small improvements and it is not a new booking system. When a request is genuinely a project, a good supplier says so and quotes it, and a supplier who tries to squeeze it into the retainer will deliver it badly.
The plans that work have both, named separately: a fixed amount for keeping it alive, which you should think of as insurance and not resent, and a stated allowance for making it better, which you should actively spend. If you have never once used your improvement hours, you are on the wrong plan or with the wrong supplier, and either is worth a conversation.
How to compare two plans
- Ask what happens if you stop paying. Everything else is secondary to this answer.
- Ask for last month's report from an existing client, redacted. Real work leaves a trace; if nothing can be shown, ask what you would be paying for.
- Separate response from repair, in writing, with hours of cover stated.
- Ask what is excluded. A good supplier answers immediately and specifically, because they have thought about it.
- Ask when a restore was last tested. The answer tells you whether backups are a process or a checkbox.
- Ask who you speak to, and whether it is the same person each time.
- Compare against the alternative honestly: doing nothing and paying for repairs as they arise. For a genuinely simple site that is sometimes the better deal, and a supplier worth trusting will say so.
In our own case those pieces are split deliberately so you can buy only what you need: security updates, monitoring, support with stated response times, and continuous improvement for the hours that actually make the site better rather than merely keeping it alive.
Not for you if
One test if you are unsure whether your current plan is worth it: ask what was done last month. If there is a clear answer, you are buying maintenance. If the question is awkward, you have found out something useful for the cost of an email.