The first process to automate should always be the one that hurts the most – the task you do most often, the one that eats the most time, the one you’d most like to never touch again. Three questions determine whether something is worth automating: do you do it at least three times a week? Is it the same process every time? Can it be fully described and regulated, with no individual decision-making required? If all three are yes, automate it.

The example is a nutrition consultant client: paper-based or document assessments that clients filled out slowly or incompletely were replaced by a simple form connected to a Make webhook. The data goes to ChatGPT, which analyses it based on pre-defined criteria, a JSON module processes the output, and PDFShift converts everything into a professional PDF proposal – automatically, with no manual layout or editing. All responses are also saved to a Google Sheet for ongoing statistics. The entire flow is six modules.

Closing advice: don’t start with a 45-module sales funnel. Start with something simple like a quote, learn from it, keep refining it, and it matures on its own over a few months. By then, the next step – a larger, more complex process – is a natural progression.

This was just one example – and six modules. In the full course, we build eight of these, from a 3-module beginner flow all the way to a 45-module sales funnel that qualifies leads, sends personalised offers, and follows up while you sleep.

If you’re ready to stop doing manually what a machine can do better – I’ll see you inside.