The most common question I get about fractional engineering is also the most reasonable one: what am I actually paying for each month? Retainers have a reputation problem, earned by every agency that ever invoiced monthly for work nobody could point to. So here is an ordinary retainer month in my practice, in the order it actually happens. Nothing glamorous appears below, which is precisely the point.
The invisible layer: monitoring that never sleeps
Underneath everything, monitoring runs continuously: uptime checks, performance baselines, security signals, backup verification. Most months this layer produces nothing but quiet confirmation. The month it produces an alert instead, you find out about the problem from me, with a diagnosis attached, rather than from a customer or a blank screen. This is the part of the retainer you cannot see and would miss most.
Week by week: the maintenance rhythm
Every month has a maintenance pass: security updates applied, dependencies brought current, backups tested rather than assumed, and the small degradations that accumulate in any live system caught early. A database that has grown sluggish. A plugin update that needs review before it can be trusted. A disk creeping toward full. Individually trivial; collectively the difference between a system that lasts and one that quietly rots.
The visible layer: the improvements queue
On the Partner and Embedded tiers, development hours are part of the month. Clients keep a running queue, and we work it in priority order: a new automation, a landing page, a performance push, a data report the business suddenly needs. Because the same engineer runs the whole system, there is no onboarding tax on any of it. The person building your new automation already knows why your server is configured the way it is, because he configured it.
The report: what you actually receive
Every month closes with a report in plain language: what was monitored and what it showed, what was updated, what was built, what I recommend next, and anything that needs a decision from you. No jargon, no padding. If a month was quiet, the report says it was quiet and why that is good news. You should never have to wonder what a retainer did for you; the answer should arrive in writing.
When something breaks anyway
Systems fail; the retainer determines what failure costs you. Incident response is part of every tier: I find out from the monitoring, usually before you do, and the person who answers is the person who built the thing, which shortens every diagnosis. The response-time commitments tighten as tiers go up, and Partner and Embedded clients get priority when two things happen at once, which is a real scenario a three-client cap keeps rare.
The terms, plainly
Retainers run from €1,200 a month for the monitoring-and-maintenance layer, €2,500 with development hours, and €5,000 fully embedded. Three month minimum to reach steady state, then rolling monthly with 30 days notice. No annual lock-in, because the honest retention mechanism is that walking away from someone who already understands your systems makes no sense. The full tier breakdown is on the pricing page.
If your current provider invoices monthly and you could not describe what last month contained, that gap is worth examining. Bring it to a free 45-minute diagnostic call, and you will leave knowing what a properly run month should look like for your systems, whoever ends up running it.