You will be able to explain the main design differences between Zapier, Make and n8n and which suits which builder.
Mei Ling's enquiry flow from module 2 has one filter, one branch by level and three actions. If you picture that same flow built three times, once each in Zapier, Make and n8n, you have a fair way to compare the three tools.
Ask three people which automation tool to use and you'll likely get three confident answers, each naming a different tool. People tend to recommend the tool they learned first. Online comparisons aren't much better, because they tend to recommend whichever tool pays the writer. Each tool's design gives you a better basis for choosing, since design decides what feels easy and what feels like a fight. All three tools can run Mei Ling's flow, but each asks you to think about it differently. These products change often, so treat what follows as each tool's general approach, and check each tool's current pages for features, plans and prices before you commit.
In Zapier, a flow is a vertical list. The trigger sits at the top and each action is a numbered step below it, so you read the flow from top to bottom, like a recipe. Branching exists, usually as separate paths that each continue as their own list, but the basic shape is still a straight line.
That shape suits mostly linear flows, and most small-business flows are linear: one event, then three or four things follow. Mei Ling's enquiry flow is mostly linear, so it would be quick to build in Zapier. Zapier also connects to a very large catalogue of apps, so an obscure or local app has a reasonable chance of being listed. Check rather than assume.
Zapier gets awkward with heavy branching and with lists of items. A flow with five branches, each with its own sub-branches, becomes a long scroll that is hard to see in one view.
Make puts a flow on a visual canvas. Each step is a circle, the circles are joined by lines, and you can see the whole flow at once. You build branches with a router, which splits the line into several paths you can see side by side. You handle lists with an iterator, which takes a list of items and sends each one through the following steps in turn. If an order has five lines, the iterator sends each of the five through the steps after it. This is the loop from Lesson 2.2, "Data mapping: passing fields from one step to the next".
Make suits flows with many branches and a lot of data handling, such as splitting an order into lines, reformatting each line and writing the lines to different sheets. The canvas also makes a complex flow easier to explain to someone else, because they can see its shape. The downside is a slightly steeper start. Each step shows more settings, and your first hour can feel busier than it would with a simple list.
n8n also uses a visual layout, with each step as a node joined to the next. Where it runs is what sets it apart. You can use it as a hosted service, or install it on your own server, a rented cloud machine or an office computer. Running it on your own machine instead of the vendor's service is called self-hosting. It keeps the data that passes through your flows on a machine you control, which some organisations prefer. Lesson 3.3, "What a tool can see once you connect your accounts", covers what self-hosting changes for data protection. Self-hosting takes work: installing, updating, securing and making sure it keeps running. It suits people who are comfortable with a little technical setup, or people with an IT colleague willing to help. n8n also lets you add small pieces of code inside a flow if you want to, though you never have to.
Each tool bills by its own usage unit, meaning the thing it counts for billing. Some tools count every action step that runs, so a flow with five steps uses five units each time it fires. Others count each whole run of a workflow, however many steps it has. Whether filters, branches and failed runs count also depends on the tool, so the same flow can be cheap in one tool and expensive in another. A flow with many steps that runs a few times a day costs very differently from a two-step flow that runs a thousand times a day.
Don't compare plan prices on their own. Estimate how many runs a month your flow will have and how many steps each run uses. Then read each tool's current pricing page to see how it would count those runs and steps, including whether filters, branches and failed runs count, and compare those results. Module 8 comes back to usage-based cost when you build a cost sheet.
Start by confirming that the tool connects to your apps, with the triggers and actions you need. Check this before anything else. Next, look at your flow's shape. It might be mostly linear, heavily branched, list-heavy, or carrying data that must stay on your own server. Match that shape to a tool's design using the rough guide below, and give your own experience with a tool more weight than any guide.
Mostly straight-line flows between popular apps suit a list-style tool like Zapier, which is often quickest. Flows with many branches or lists of items suit a canvas tool like Make, which is often clearer. If data should stay on your own server and someone is willing to run it, self-hosted n8n is worth a look.
Before you commit to a tool, check its current pages for features, plans and prices.
For the activity below, take the two apps in your chosen workflow. In each of Zapier, Make and n8n, check that both apps connect with the triggers and actions you need, and note anything missing.
Look up whether the two apps in your chosen workflow are supported in each of the three tools and note any missing trigger or action.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).