Made for Swiss self-employed & SMEs
Every week brings a new AI tool, a fresh set of promises, and another reason to feel you are falling behind. But for a Swiss SME or a self-employed professional, starting with AI is not about chasing the newest model. It is about picking one boring, repetitive task and doing it a little better — then repeating.
The businesses that get real value from AI rarely begin with an “AI strategy” or a costly rollout. They begin with a single inbox, a single template, or a single hour of typing they want back.
Start with the work, not the tool
The most common mistake is to start by shopping. People compare ChatGPT, Claude, Copilot and a dozen agents, read reviews, sign up for three trials — and never actually automate anything.
Turn it around. For one week, write down where your hours actually go. Which task repeats every day or every week, takes ten minutes or more, and needs no deep judgment? That is your first candidate.
Three tasks that usually pay off first
In our work with Swiss freelancers, notaries, doctors and small offices, three tasks almost always repay the effort:
- Drafting. Emails, client letters, offers, reminders. AI produces a solid first version in your tone; you review and send.
- Summarising. Meeting notes, contracts, official letters. AI turns a long document into a short brief with the key points.
- Translating. Writing to clients in German, French, Italian or English without starting from scratch.
Start with exactly one of these. One task, one channel, one clear win.
Which tool is actually enough
For a first workflow, a general assistant such as ChatGPT or Claude is enough. You do not need AI agents, and you do not need to run a model on your own server on day one. A single subscription, around CHF 20–40 per month, covers the drafting, summarising and translating above. You can compare the real costs in our separate guide before you commit.
The data rule that keeps you safe
Before you paste anything into a tool, one rule matters: treat client data as confidential, even in a friendly chatbot. Under the Swiss Data Protection Act (DSG), do not feed identifiable client information — names, medical notes, financial details — into a public consumer tool without checking.
The practical fix is simple: anonymise or replace identifying details first, and for sensitive work in medicine, law or finance, use an approved or local setup. The goal is to keep the habit useful without ever compromising confidentiality.
A 30-day starter plan
- Week 1: Note where your time goes and pick the single task.
- Week 2: Write one good prompt or template for that task, using your own examples.
- Week 3: Use it every time the task comes up, and adjust the output until it sounds like you.
- Week 4: Decide: keep it, improve it, or drop it. Then pick the next task.
That is the honest version of “getting started.” No transformation, no agents, no hype — just one repetitive task that no longer eats your afternoon.
When to level up
Once one routine runs reliably and saves you real time every week, then it makes sense to explore more: automation across several steps, or a local model when client data must never leave your machine. We have written about both. But do that second — after you have proof that the first step was worth it.
