Monitoring as a default companion: the agent-run alternative to a retainer
Recurring revenue doesn't have to mean standing claim on senior time. Monitoring + updates is the leveraged version.
Retainers are simple to sell and miserable to scale. You commit hours; the client commits dollars; both of you renegotiate the value once a quarter. The leverage caps at the number of senior people you can afford.
What monitoring is, specifically
Agents watch the system you shipped to the client. When something drifts — schema validation fails, lighthouse drops below the threshold, an integration starts rate-limiting — the agent posts a recommendation to the client's channel. Including how much it would cost to fix.
Updates means the client gets every improvement we ship to the underlying product. Their deployment stays current without us touching it.
What it isn't
Monitoring is not free fixing. It's monitoring plus a recommended fix, where doing the fix is paid Support — metered, hourly, with a 10-hour pack at a discount, no refund on unused, expires.
That distinction sounds petty until you've shipped fifteen systems and the support load starts eating senior time. The honest version: every hour of 'free monitoring' fixes is an hour you can't spend shipping the next system. Multiplied by fifteen clients, you've built an agency.
Why this works as recurring revenue
At twenty clients on monitoring, the recurring base is real money — and it doesn't grow with senior hours. The base scales with the number of systems running, which scales with the number of setups sold, which is the loop we want compounding.