Zapier vs Make vs n8n: Which Automation Tool Fits Your VA Workflows
When a founder hires a virtual assistant through us and the work settles into a rhythm, the same question usually turns up around week six: "We keep doing this by hand. Can we automate it?" The answer is usually yes, and the next question is which tool. Nine times out of ten the shortlist is Zapier, Make and n8n.
I run Virtual Assistant Capital, and I don't have a favourite among the three. The VAs we place have built workflows in all of them, and each one has a business it suits and a business it frustrates. This guide is the comparison I give founders when they ask: what each tool is, how each one charges, who should use which, and the part most comparisons skip, which is who is going to own the automation once it is built.
I am not going to quote plan prices here, because all three change their pricing often and you should read the current numbers on their own pages. I will describe how each one counts usage, because that is what decides your bill, and link to the source.
The short answer first
- Zapier is the easiest to start with and connects to the most apps. It is the right default if the person building the workflows is not technical and the workflows are simple.
- Make gives you a visual canvas with more control over branching, loops and data handling, and its usage model often stretches further for multi-step workflows. It suits a VA who is comfortable thinking in flowcharts.
- n8n is the most flexible and the only one of the three you can run on your own server. It suits a business with someone technical on hand, high volumes, or data it would rather keep in its own infrastructure.
If that is all you needed, the rest of this guide explains why, and how to decide when the answer is not obvious.
What these tools actually do
All three are workflow automation platforms. You choose a trigger, something that happens in one app, and then one or more actions that should follow in other apps. A form is submitted, so a contact is created in the CRM, a welcome email goes out and a task is assigned to your VA. An invoice is paid, so a spreadsheet row is updated and a thank-you note is sent.
The work they replace is the copying and pasting between tools that eats a surprising share of a small business's week. They are not AI agents and they are not a replacement for a person. They do exactly what they are told, every time, including the wrong thing if the process was set up wrong. We go deeper into that distinction in virtual assistant vs AI.
How each one charges, and why it matters
The headline price of an automation tool tells you very little. What matters is the unit it counts, because the same workflow can cost very differently on each platform.
- Zapier counts tasks. According to Zapier's pricing page, a task is counted whenever Zapier successfully completes a unit of work for you, which in practice means each action step that runs. Triggers do not count, failed steps do not count, and built-in tools such as Filter, Formatter, Paths and Delay are listed as non-billable. The free plan includes 100 tasks a month and is limited to two-step Zaps: one trigger and one action. Multi-step Zaps need a paid plan.
- Make counts credits. Make's pricing page says credits became its billing unit on August 27, replacing the older operations model. Each action usually uses one credit, and some AI features can use more. The free plan includes 1,000 credits a month.
- n8n counts executions. n8n's pricing page says pricing is based on monthly workflow executions "regardless of complexity". One execution is one full run of a workflow, however many steps it has. There is no free-forever cloud plan, but there is a cloud trial, and n8n publishes a self-hosted version on GitHub.
Here is why the unit matters. Imagine a workflow with eight steps that runs 500 times a month. On a per-step model, that is around 4,000 billable units. On a per-run model, it is 500. The longer and busier your workflows get, the more the counting method shapes your bill. For a handful of simple two-step workflows, the difference is small and ease of use matters more.
Zapier: the easiest place to start
Zapier is where most of the founders we work with already have an account, often one they set up themselves years ago and half forgot about.
Where it shines:
- Ease of use. The builder walks you through each step in plain language. A VA who has never built an automation can usually get a simple workflow running the same day.
- App coverage. Zapier connects to a very large number of apps, and niche industry tools are more likely to have a Zapier integration than anything else. If your CRM, scheduling tool or practice software is obscure, check Zapier first.
- Reliability and support documentation. Common problems have been solved and written up many times.
Where it frustrates:
- Cost at volume. Because each action step counts, long or high-frequency workflows add up faster than on the other two.
- Complex logic. Paths, filters and loops are possible, but complicated branching gets harder to read and maintain than on a visual canvas.
- The free plan is narrow. Two-step Zaps cover very simple handoffs. Most real business workflows need more steps.
Best fit: a small business with a non-technical owner or VA, a modest number of workflows, and apps that are not all mainstream.
Make: more control for a modest learning curve
Make lays out each workflow, which it calls a scenario, as a visual map of connected modules. You can see the whole flow at once, including branches, error routes and loops.
Where it shines:
- Seeing the whole workflow. For anything with more than a few steps, the visual layout makes it much easier to understand what happens and where it broke.
- Data handling. Make is good at working with lists, transforming data between formats and processing many records in one run, which matters for things like syncing orders or cleaning up spreadsheets.
- Usage that stretches. For many multi-step workflows, the credit allowance goes further than a comparable per-task allowance would. Check against your own workflows, because this depends heavily on how they are built.
Where it frustrates:
- A steeper start. The canvas is powerful, but a VA new to automation will need a few days to get comfortable with it, and more for advanced features.
- Fewer niche integrations than Zapier. Most mainstream apps are there. For less common ones, you may need to use a generic HTTP module, which means reading an API's documentation.
- Easy to overbuild. Because the canvas makes complexity possible, workflows sometimes grow into something only the builder understands. That is a management problem more than a tool problem, and I come back to it below.
Best fit: a business whose workflows have real logic in them, with a VA who likes systems and is willing to learn the tool properly.
n8n: maximum flexibility, more responsibility
n8n is built with technical users in mind. It also uses a visual node editor, but it lets you drop in code where you need it, call almost any API, and run the whole thing on your own server.
Where it shines:
- Flexibility. If a workflow needs custom logic, n8n rarely says no. You can write JavaScript or Python inside a step.
- Self-hosting. You can run n8n on infrastructure you control. For businesses careful about where client data travels, that is a meaningful option, not a gimmick. Read the license terms on n8n's site before you rely on it for your business.
- Predictable cost for complex workflows. Because billing on its cloud plans is per execution rather than per step, long workflows do not get more expensive just for being long.
- AI workflows. n8n has put a lot of effort into AI building blocks, so it is a common choice for workflows that combine automation with AI models.
Where it frustrates:
- It assumes technical comfort. A non-technical VA can learn the basics, but the power is in the parts that look like programming.
- Self-hosting is a job. If you host it yourself, someone owns updates, backups, security and uptime. "Free to self-host" means free of a subscription, not free of work.
- Smaller integration library. Fewer ready-made connectors than Zapier, offset by the ease of calling APIs directly.
Best fit: a business with technical help available, high workflow volume, data it wants to keep in-house, or ambitions to build AI-heavy workflows.
Side-by-side comparison
| Zapier | Make | n8n | |
|---|---|---|---|
| Billing unit | Tasks (completed action steps) | Credits (usually 1 per action) | Workflow executions (one per full run) |
| Free option | 100 tasks a month, two-step Zaps | 1,000 credits a month | Cloud trial; self-hosted version on GitHub |
| Learning curve | Lowest | Moderate | Highest |
| Ready-made integrations | Largest library | Large | Smaller, strong API and code support |
| Complex logic | Possible, gets hard to read | Strong, visual | Strongest, including code |
| Self-hosting | No | No | Yes |
| Best for | Non-technical teams, simple workflows | Logic-heavy workflows, systems-minded VAs | Technical teams, volume, data control |
Free-plan details are taken from each company's pricing page, linked above, as of this writing. Check them again before you choose, because they change.
The question that matters more than the tool
Here is the part I spend most time on with founders. The tool is rarely why an automation project fails. Ownership is.
Every automation breaks eventually. An app changes its fields, a password expires, a form gets a new question, a customer types their phone number in the email box. When that happens, one of two things follows. Either someone notices, fixes it and updates the documentation, or the workflow quietly stops working and you find out weeks later when a client asks why nobody replied.
So before choosing a tool, answer these questions:
- Who builds it? If it is you, choose the tool you will actually learn. If it is your VA, choose the tool that matches their skills, or budget time for them to learn.
- Who gets the error alerts? Every platform can notify someone when a run fails. That someone should be a person with the time and access to fix it, not an inbox nobody reads.
- Where is it documented? Each workflow should have a short written note: what triggers it, what it does, which accounts it uses, and what to do if it breaks. Without that, the automation belongs to whoever built it and leaves when they do.
- What stays manual on purpose? Anything that sends money, commits you to a client, or speaks for you in a sensitive situation should have a human approval step. Automate the preparation, not the decision.
- Whose accounts does it run on? Workflows should be built in accounts the business owns, with the VA given access, never in a personal account. Our VA security and NDA guide covers how to set up access properly.
This is why we built the AI Manager plan around a full-time, AI-fluent VA: the value is less in building the first workflow and more in running, checking and maintaining all of them as the business changes.
How a VA usually rolls out automation
When one of our VAs automates part of a client's business, the order is almost always the same, whatever the tool:
- Run the process by hand first. For the first few weeks, the VA does the work manually and writes down exactly how it runs. This is also part of a good first week of onboarding.
- Pick one painful, repetitive process. Not the most complex one. The one that wastes the most time each week with the fewest exceptions.
- Build it with a human checkpoint. The first version usually prepares the work and lets the VA approve it before anything goes out.
- Test it with real examples. Including the awkward ones: missing fields, duplicates, unusual formats.
- Remove the checkpoint only where it is safe. Internal updates can usually run unattended. Client-facing messages often keep a check for good.
- Review monthly. What failed, what changed, what new process is worth automating next.
Starting this way also makes the tool choice easier. Once you have one working workflow and know what the rest of your list looks like, it becomes obvious whether you need Zapier's integrations, Make's logic or n8n's flexibility.
Common mistakes
- Choosing on the free plan alone. Free tiers are designed to show you the product, not run your business. Price the plan you will need in three months.
- Automating a messy process. If the manual process has no clear rules, the automation will copy the mess faster. Write the process down first.
- Running three platforms at once. It happens by accident: a Zapier account from years ago, a Make scenario a freelancer built, an n8n experiment. Pick a primary tool and move workflows over as they need attention.
- No owner for failures. The single biggest cause of automation problems we see. Every workflow needs a named person who gets the alert.
- Letting automation talk to clients unchecked. A wrong automated message to an important client costs more than any subscription saves.
FAQ
Is Zapier, Make or n8n best for a small business? For most small businesses with non-technical staff and simple workflows, Zapier is the easiest start. If your workflows have a lot of branching or data handling, Make is often a better fit. If you have technical help and want self-hosting or very high volumes, look at n8n.
Which is cheapest: Zapier, Make or n8n? It depends on how your workflows are built. Zapier counts completed action steps, Make counts credits per action, and n8n counts whole workflow runs. Long, busy workflows tend to cost less on a per-run model; a few simple workflows cost little anywhere. Price your actual workflows on each company's pricing page.
Can a virtual assistant build automations in these tools? Yes. Many VAs build and maintain workflows in Zapier and Make, and some are comfortable in n8n. Ask for an example of a workflow they built and how they handled it when it broke. Our workflow automation VA page explains what that role covers.
Do I need to know how to code to use n8n? Not for basic workflows, but n8n is at its best when someone can write small pieces of code and work with APIs. If nobody on your team is comfortable with that, Zapier or Make will usually be faster to get value from.
Can I switch from Zapier to Make or n8n later? Yes, but there is no automatic migration: each workflow has to be rebuilt and tested. That is a good reason to keep a short written description of every workflow, so a rebuild is a few hours of work rather than detective work.
Which plan should I choose if I want a VA to run my automations? For light automation alongside other work, our Executive plan at $1,599 a month can scope basic Zapier or Make connections. If automation and AI-assisted workflows are a core part of the job, the AI Manager plan at $2,599 a month is built for it. See pricing for what each plan includes.
The short version
Zapier is the easiest and connects to the most apps. Make gives you more control over complex logic and often stretches further on multi-step workflows. n8n is the most flexible, bills per run and can be self-hosted, but it expects technical comfort. Choose by who will build and maintain the workflows, not by the feature list. Whichever tool you pick, give every workflow a named owner, written documentation and a human check wherever a mistake would reach a client or your money.
Get a free VA shortlist in 72 hours.
Tell us what you need. We send 2–3 hand-picked, vetted VAs to interview. $0 recruitment fee, and you pay nothing until your VA starts. Essentials $899/mo, Executive $1,599/mo, AI Manager $2,599/mo, month-to-month.