The maintenance work agencies do free inside retainers

Most agency retainers are sold as strategy, design or development time. Then a client's site breaks, and fixing it comes out of the retainer — because saying no to a retained client over a broken site is not a conversation anyone wants.

Over a year that adds up to a meaningful part of what you sold as creative capacity.

Where it goes

The unplanned fixes. A plugin update breaks a layout. A form stops sending. The site is slow this week. Each is twenty to ninety minutes, and each arrives without warning, usually mid-project.

The small requests. Change the phone number, swap a photo, add a team member, update the opening hours. Five minutes each, except that switching context into a site you have not touched for six weeks is not five minutes.

The reassurance. "Is our site secure?" "Should we be worried about this email from our host?" Not billable, not avoidable.

The emergencies. A hacked site consumes a week of somebody's attention, and it is nobody's planned work.

Working out the cost

Look at last quarter's time logs and pull out everything that was maintenance rather than the work the retainer was sold for. If you do not track it, estimate: 30–60 minutes per site per month, plus the ad-hoc requests.

Ten retained clients at an hour a month is 120 hours a year. At $80 billable, that is nearly $10,000 of capacity — spent on work you did not price, cannot forecast, and that interrupts the work you did price.

And the interruption costs more than the hours. A developer pulled out of a build loses the twenty-minute fix plus the time to get back into what they were doing.

Why it persists

Three reasons, all understandable:

Saying no feels worse than absorbing it. Their site is down and you built it.

It is invisible individually. No single request is worth a conversation. It is only the aggregate that hurts, and nobody looks at the aggregate.

There is no alternative on the table. If you have no care plan to point at, the only options are do it free or refuse.

That third one is the actual fix.

Separating it

1. Measure it for a month. Tag maintenance time separately. You need a real number before the conversation, because "it feels like a lot" is not persuasive to you or to them.

2. Define the line. Maintenance is keeping the existing site working: updates, backups, security, fixing what breaks, small edits to existing content. Retainer work is making the site do something new. Examples matter more than definitions.

3. Introduce a care plan as an addition, not a change. The framing that works is that they have been getting this informally, with no guaranteed response time, and you want to make it proper. More on that conversation here.

4. Price it honestly. $100–200 a month for a normal site. It should look small next to the retainer, because it is.

5. Protect the retainer hours. Once maintenance is its own line, retainer time goes back to being what it was sold as — and you can point at the plan rather than refusing.

What changes

Revenue that is predictable rather than project-shaped. Developers who are not interrupted. And a clear answer when a client asks whether their site is being looked after — which, at most agencies, currently has no honest answer.

The delivery is the part that stops people, since adding maintenance means adding the work. White label removes that: you sell the plan, keep the relationship and the margin, and someone else runs the weekly work under your name. At $26–39 a site against $150 charged, it turns a cost centre you were absorbing into a line that pays.