How to choose a Make automation niche

Short answer

Choose one process and one type of business for a short test. Look for repeated data entry, lead routing, reporting, onboarding, or order updates. Talk to people who do the work before building a large offer. Keep the niche if you find the same problem and a real next step. Change it if you only collect compliments.

Start with a process, not a sector

"Real estate" is not an offer. "New enquiry to CRM, owner alert, and follow-up task" is a process. "Online shops" is broad. "Paid order to stock alert and internal handoff" is a process.

Make's freelancer guide tells new freelancers to list skills, tools, industries, and processes, then check demand. That order is sensible. Your tool knowledge is useful only when it meets a process someone wants changed.

Write down five processes you have seen at work or in a business you understand. For each, note who performs it, what starts it, what they copy, where errors occur, and what happens when it is missed. If you cannot describe the current manual path, you are not ready to sell an automation for it.

Find signs of active demand

Look for evidence that the work exists now:

• a job post for an assistant handling repetitive admin
• a business owner asking how to connect two systems
• a support thread about duplicate records or missed notifications
• a process document with several manual handoffs
• a freelancer or agency turning away the same type of request

These are signals, not proof that a buyer will pay you. Ask a factual question about the past: "How did you handle the last batch of enquiries?" or "How often does someone copy this into the CRM?" Do not begin by explaining your solution.

ToolNav's Make consultancy page recommends targeting small businesses doing obviously manual work, and lists real estate agencies, e-commerce brands and marketing agencies as niches that "pay consistently". It also suggests asking an AI chatbot to estimate hours saved and what clients would pay. Those are the page's claims and an estimate from a chatbot, not market data, so this guide does not repeat them. The page's advice to pick a niche you already know something about, and to build one demo, is a reasonable test.

Build one demo with a boundary

A good demo has one trigger, one useful path, and a visible handoff. It may take a form submission, check a field, create a record, and alert a person. Make the boundary clear. The demo is not a promise to automate every process in that industry.

Use test accounts and sample records. Keep a note of what would need client access. Add a failed-input path and say what the person sees when something goes wrong. The error handling guide for client automations covers the next layer.

Record a short walkthrough or publish a simple page. Explain the problem in the buyer's language first. Put the module names later, if they matter.

Test the niche with conversations

Speak to people who perform the process. Ask:

• When did you last do this manually?
• What information did you move?
• Which part was slow or easy to get wrong?
• What did you try already?
• Who would approve a change?

Do not ask, "Would you use my automation?" People are polite about ideas. Past behavior is better evidence. If they have never looked for a fix, ask what they do when the task becomes too large.

A positive answer is not yet a lead. A stronger signal is a person sharing a real example, showing the current setup, introducing the decision maker, or agreeing to a scoped review.

Turn the pattern into a small offer

After several conversations, write the offer as a process and boundary:

"I set up a Make workflow that sends new website enquiries to your CRM and alerts the assigned person. The first version covers one form, one CRM, and one alert route."

Do not promise hours saved unless the client has measured the current work and you have a defensible way to compare it. State what you will test after setup.

Use a fixed scope when the systems and path are known. Use a paid discovery step when the current process is unclear. For pricing models and client-owned accounts, read how to price n8n and Make automation work.

How to choose between two niches

Score each candidate in a small notebook. Record whether you found a repeated process, a person who owns it, access to a decision maker, a working demo, and a next conversation. Do not turn the score into fake precision. The notes are there so you remember what actually happened.

Keep the niche that produces specific examples and next steps. Drop the one that produces only broad interest. Revisit it after you have tested another process. A short test teaches more than an elaborate site aimed at an imaginary buyer.

Find clients without making the offer broad

Start with people who have already mentioned the process, agencies that serve the business type, and communities where the question is allowed. The existing guide on first automation clients without Upwork covers channels and conduct. Follow community rules and check the laws that apply before cold outreach.

Keep a log with the niche, process, source, question asked, reply, and next step. If the same problem appears repeatedly, update the demo. If the replies are all about a different process, change the offer rather than forcing the first one.

Thomas Partners brings client work to freelancers. You keep 90% of the profit after documented costs are repaid. Apply to join.