Every agency we've ever competed with bills by the hour. They will tell you it's the "fair" model — you pay for exactly what you get. But hourly billing has a structural problem that nobody talks about honestly: it rewards slowness and punishes expertise.
The better you get at building AI systems, the faster you build them. If you're billing by the hour, that means your income drops as your skills improve. The perverse incentive is to stay slow, to pad scopes, to invoice for discovery that could take two hours instead of eight.
Hourly billing punishes you for getting good. The better you get, the less you earn.
We built Fortis to avoid that entirely. We charge for the system — and the ongoing value of having us maintain it — not for the hours it took to ship it.
What we actually charge for
Our pricing has two components. The first is the setup fee — scoped after your discovery call based on the complexity of what needs to be built. We're transparent about what goes into this number: the build hours, the testing, the integration work. This is a one-time cost, and we quote it before you commit to anything.
The second is the monthly retainer. This is what keeps the system maintained, updated, and working as the tools underneath it evolve. AI models update. APIs change. n8n releases new nodes. A system you deploy today without maintenance will drift over six to twelve months. The retainer is what prevents that.
Critically, the retainer is not a lock-in. If you cancel, the system keeps running. We just stop actively maintaining it. You own what we built. We don't hold it hostage.
The 12-month maths
Let's run a simple comparison. Say you engage a boutique agency to build an AI receptionist on an hourly model — $200/hr is common in the Melbourne market. A project like the Bodyline build would be quoted at roughly 60–80 hours. That's $12,000–$16,000 up front. Then, because they've billed and moved on, you're on your own when something breaks or the model needs updating.
With Fortis, the setup fee is scoped and transparent. The ongoing retainer from $650/mo means that across 12 months, your total cost is known in advance — and includes everything staying live, maintained, and improved. No surprise invoices for an extra sprint. No hourly debate about whether something was in scope.
Over 12 months, clients typically find that the predictable monthly cost is similar to or lower than what they'd have spent hourly — and they get a maintained, working system at the end of it rather than a codebase they can't touch.
Why this is better for both sides
This model changes the incentive structure completely. When we're on a retainer, we want the system to work. A broken voice agent that generates complaint calls for our client is a client we'll lose. Our incentive is entirely aligned with the system performing well — not with logging more hours.
It also means we're honest about scope early. If something doesn't need automation yet, we'll say so. We've turned away builds where the process was too immature or the volume too low to justify it. Selling someone a system they don't need is how you build a business that churns.
Hourly billing creates a consultancy that's always looking for the next engagement. Retainer billing creates one that's invested in keeping you.
That's the model we've built. It's not the easiest one to sell — most clients have been conditioned to expect an hourly quote. But it's the honest one.