YouTube is full of “AI automation agency” videos that make this sound like passive income. It isn’t. Make.com is a genuinely good tool for getting paid to connect other people’s software, but the money comes from understanding a business’s tangle of forms, spreadsheets and inboxes, not from the tool itself. Here’s a plan that starts small and grows with your skill.
The short version
What this takes
| Best method | Fixing one annoying process for small businesses |
| Cost to learn | $0: the free plan plus the free Make Academy |
| Client’s tool cost | From $12 a month, paid by the client on their own account |
| Going rate | $35 to $60 an hour for AI automation work, per Upwork’s hiring guide |
| Income | Not guaranteed |
If you haven’t used Make yet, start with my Make.com review. It explains the credit system, which matters a lot when you quote clients.
Three ways to earn with Make.com
| Method | Who pays you | What it can pay | Effort |
|---|---|---|---|
| Automation projects for small businesses | Business owners | Your project or hourly fee, plus a monthly care plan | Medium to high |
| Selling scenario templates | Other Make users | Small amounts per sale | Low to medium |
| Affiliate program | Make | 35% of a referral’s payments for 12 months | Needs an audience |
Method 1: Fix one process for small businesses
Forget “AI employees”. The work small businesses will actually pay for is dull. New website enquiries get copied into a spreadsheet by hand. Invoices go out late. Appointment reminders depend on someone remembering. Review requests never get sent. Each of these is a Make scenario of a few modules, and each saves the owner a few hours a week, which is an easy thing to put a price on.
Learn it properly first
Make Academy is free. It comes with digital badges and a final assessment for certification, and all it needs is a free Make account. Do the courses and build three automations for yourself or a friend’s business before you charge anyone. Your first client shouldn’t be the one paying for your learning curve.
Build in the client’s account, not yours
Have the client open their own Make account and pay Make directly, then invite you to it. They own their data and their automations, and if you part ways nothing breaks. If you build a scenario on your side first, you can export it as a blueprint, a JSON file under 2 MB. Blueprints don’t include connections, though, so the client still has to connect their own apps after import.
Price the result, then the upkeep
Upwork’s own hiring guide says AI automation engineers generally cost $35 to $60 an hour, and that project-based pricing is common. Cheaper offers exist: Make automation gigs on Fiverr start from about $30. Quote a fixed price for the build, then offer a small monthly care plan to watch for errors, fix breakages when an app changes, and make small tweaks. Apps do change, so the care plan is not busywork.
Worked example
A local clinic gets about 30 website enquiries a day. You build a scenario that catches each enquiry instantly through a webhook, adds it to their spreadsheet, writes a short summary, and alerts the front desk: about six credits per enquiry. That’s around 900 enquiries and 5,400 credits a month, which fits inside the $12 Core plan’s 10,000 credits. The client pays Make directly.
You charge 10 hours at $45, which is $450, plus a $75 monthly care plan. Four clients like this would bring $1,800 in setup fees and $300 a month. These are illustrative prices you’d set yourself, and it takes time to reach four clients.
Where the first clients come from
Start with businesses you already know: your dentist, your gym, a friend’s shop. Ask what job they hate doing every week. Upwork has a steady stream of Make.com job posts once you have a portfolio of real builds. Later, if you grow into a small team with certified experts, the Make Solution Partner program is free for qualifying agencies. It requires partner training certifications, and only its Silver, Gold and Platinum tiers receive leads, so treat it as a later step.
Method 2: Sell templates other people can import
Once you’ve built something useful, like a content calendar pipeline or a lead-scoring flow, you can export it as a blueprint and sell it from your own site. Since blueprints carry no connections, each buyer connects their own apps, so include a short setup video and an honest estimate of how many credits the scenario uses a month. Expect small money per sale. Templates work best as proof of skill that brings in project clients.
Method 3: The Make affiliate program
Make pays affiliates 35% of a referral’s subscription payments for 12 months. Someone who signs up through your link has 30 days to do it. Four details most write-ups skip:
- The 12 months start at registration, not at the first payment. If someone signs up free and upgrades eight months later, you earn for four months.
- No payout until you have $100 and three paying referrals. Payments go only through Wise.
- Extra credit purchases don’t count. Commission is earned only on subscription payments.
- No “Make” in ads. You can’t use the trademark anywhere in paid ad copy without written consent.
On monthly prices, a Core subscriber who pays for a full year is worth $50.40 to you, a Pro subscriber $88.20, and a Teams subscriber $159.60. Your own clients are the most natural referrals, as long as you tell them you earn a commission.
Prefer tools your clients can run on their own server? n8n does the same kind of work, bills per workflow run instead of per step, and its license lets you charge to install and maintain it for clients. My guide to making money with n8n explains how that works.
Related reading: How to make money with Manychat · Systeme.io review
Mistakes that turn clients against you
- Polling triggers every minute. A trigger costs a credit every time it checks, even when there’s nothing new. One scenario checking every minute burns about 43,200 credits a month before it does anything useful. Use instant triggers and webhooks wherever the app supports them.
- Surprise AI costs. With Make’s built-in AI provider, AI steps use extra credits based on tokens. Every paid plan can instead connect the client’s own AI account, so token costs show up on a bill they can see.
- Keeping the automations on your account. It feels like job security. To the client it feels like being held hostage.
- Being careless with personal data. Agree in writing which data flows where, never send passwords or API keys by email, and point clients to Make’s data processing agreement.
- Promising AI will replace staff. Promise the specific hours saved on a specific job instead.
You’ll enjoy this if… you like puzzles, you’re patient with other people’s messy spreadsheets, and you can explain a technical thing in plain words.
It’s probably not for you if… you want to avoid client calls, or you’re hoping a tool will do the thinking for you.
The figures in this guide are worked examples, not income claims. What you earn depends on your skills, your clients and your market. This isn’t legal advice: check data-protection rules where you and your clients work. I don’t earn a commission from Make at the time of writing.
Update log: 29 September 2026: first published, with rates, prices and program terms checked that day.
Sources
- Make pricing
- Make affiliate program and affiliate help page
- Make Solution Partner program
- Make Academy
- Make: scenario blueprints
- Make: how features use credits and credits and custom AI providers
- Make: adjustments to plans and pricing
- Upwork: hire AI automation engineers
- Fiverr: Make automation gigs
- Make terms, including the data processing agreement
Leave a Reply