The underlying question
When an academy director writes to us, it's almost always the same doubt: “an AI agent sounds good, but how much will it cost me and is it really worth it for a business like mine?” It's the right question. Let's answer it with real numbers, no spin.
Before talking price, the question needs flipping. It isn't only how much the agent costs. It's how much it costs you every month not to have it: the students who write at night with no one to answer, the enrollments that go to the competitor who did reply, the hours your team spends copy-pasting the same prices. That invisible cost is usually bigger than the agent's.
What it costs, specifically
At Catalizadora we don't work with retainers or fees that keep climbing. The model is clear.
Build: a single payment
For an online academy, the entry point is MAGIA Solo: $4,500 USD, delivered in 15 days. It includes the AI agent that lives in your WhatsApp, connected to your course catalog, with lead qualification, scheduling and a payment link, plus the CRM where every conversation lands.
If your operation is larger —several locations, many programs, integrations with your current systems— there's MAGIA Core at $15,000 USD. And to build a full custom product or platform, Forge at $20,000 USD, over twelve weeks. Most academies that start out do so with Solo.
Operation: pass-through, no margin
Once live, the agent costs between $200 and $400 USD a month to run: hosting and AI tokens. That's pass-through, meaning we pass it to you at cost, with no margin on top. There's no monthly license or platform fee. Busy months use more tokens; quieter months cost less.
Compare that to the staffing alternative: an enrollment advisor costs a fixed salary every month, covers one shift and one conversation at a time. To cover 24/7 you need three shifts. The agent covers all day, every day, in parallel, for a fraction.
The “is it worth it” math
Let's run a simple calculation with numbers you can adjust to your case.
Say your average course costs $300 USD. If the agent recovers just two students a month who used to be lost for lack of a timely reply, that's $600 USD monthly that wasn't coming in before. Running the agent costs $200 to $400. From the second recovered student each month, the agent already pays for its operation and leaves profit.
And the $4,500 build amortizes fast: if the agent brings you fifteen additional students over a few months —students who would have left without it— the upfront investment is already covered. From there on, everything it captures after hours is profit that used to evaporate.
The figures change for every business, but the logic holds: in an academy that receives constant WhatsApp inquiries and loses a share for not answering, the agent pays for itself in weeks, not years.
When it is NOT worth it
We're straight about this too. If your academy gets two or three inquiries a month and you handle them fine, you don't need an agent yet. The tool makes sense when there's enough conversation volume for human support to fall short: when you lose leads to off-hours, when your team drowns in repetitive questions, when you want to grow without adding headcount.
What you definitely take with you
There's a difference worth more than the price. Under the MAGIA methodology —Mapping, Architecture, Generation, Implementation, Autonomy— the code, the data and the infrastructure end up 100% under your name. You don't rent the agent. It's yours. There's no license to expire, no platform to raise its price once you depend on it. If tomorrow you want to move everything to another provider, you take the whole thing.
That turns the spend into an investment with an asset behind it. You pay once, the asset is yours, and the monthly operation is the real cost, with no markup.
We handle your academy's search visibility separately, with a proprietary technical layer. But the math in this article is the agent that sells on WhatsApp.
Three ways to look at the return
Beyond the recovered student, there are three savings many directors don't count at first.
Your team's time. If two people spend, between them, three hours a day answering the same price and schedule questions on WhatsApp, that's sixty hours a month they're not closing enrollments or caring for students. The agent takes that repetitive load and hands those hours back to your people.
Response speed. In WhatsApp support, whoever replies first usually wins the enrollment. A human takes minutes or hours; the agent, seconds. That difference, multiplied across every inquiry, moves the conversion rate more than it seems.
Full coverage. Your academy no longer closes at night or on weekends. To your market, you're always available. That can't be bought with a salary; it's bought with an agent that doesn't sleep.
When you add up the recovered student, the freed hours and the higher conversion, the “is it worth it” question almost always answers itself for an academy with real inquiry volume.
The next step
The best way to know if it's worth it for your academy is to look at your case with your own numbers. Message our own AI agent on WhatsApp and see it live, or book a direct call with Pablo at https://cal.com/pablo-estrada-hlqaql. We'll tell you honestly whether your volume justifies the agent and, if it does, what gets built in 15 days.