n8n vs Zapier vs no automation: a 30-second test for small businesses

The tool question comes up on almost every call. Should we use Zapier? Someone mentioned n8n. Is there something better? It is the wrong first question, but it is a fair one, so here is a way to answer it in about 30 seconds without learning either platform.

Two questions decide it.

Question 1: How many steps, and how much branching?

Count the steps in the workflow, and notice how often it has to make a decision.

A short, straight-line job is a "when this happens, do that" shape. A new form submission drops a row in a sheet and sends a notification. Two or three steps, no forks. That kind of work is what Zapier and its cousins were built for, and they do it well with almost no setup.

A longer job with branches looks different. It checks a condition, goes one way or another, loops over a list, waits for a reply, then continues. Once a workflow has real logic inside it, the simple tools get expensive and awkward, and n8n (or a proper build) starts to make sense because it can hold that logic without fighting you.

Question 2: Where does the data need to go, and how sensitive is it?

The second question is about reach and control.

If everything you are connecting has a ready-made integration and the data is low-stakes, an off-the-shelf connector tool is fine. If you are touching customer records, financial data, or a system that does not have a clean integration, you want something you can host and control, where the data is not passing through a third party's servers on a plan you do not manage. n8n can run on your own infrastructure. That difference matters more as the data gets more sensitive.

The answer most of the time

Here is the honest part. For a lot of small businesses, the right first answer is "no automation yet."

Not because automation is wrong for you, but because the workflow is not ready. If the steps change every time, or the data lives in someone's head, no tool will save you. You will spend a month wiring something that breaks the first time reality does not match the recipe. A tool cannot fix a workflow that has not been decided yet. That is the "tools alone won't fix it" problem, and it is the most common thing we find.

So the 30-second test runs like this:

1. Short and straight-line, low-stakes data: a simple connector tool like Zapier is fine, and you may not need us at all.
2. Branching logic, or sensitive data, or systems without clean integrations: n8n or a custom build earns its keep.
3. Steps that change every time, or data that is not written down anywhere: automate nothing yet. Fix the process first.

A worked example

Say a 15-person business wants to stop manually copying new leads from their website into their CRM and their email tool.

Count the steps: form comes in, create the CRM record, add them to the right email list, notify the owner. Four steps, one small branch (which list depends on what they asked about). Data: contact details, not highly sensitive, but it is customer data you want handled cleanly. That points to a light automation, and if the business later wants follow-up sequences with real logic, it points toward n8n rather than stacking more connectors.

Now change one detail. Suppose the "right email list" depends on a judgment call the owner makes after a quick look. That branch is not a rule, it is a decision. The honest answer becomes: automate the intake, leave the routing to the human, and revisit once the routing rule is written down.

What to do Monday

Pick the workflow you were about to buy a tool for. Count its steps. Notice its branches. Check whether the data is written down and how sensitive it is. That is enough to know whether you need a simple connector, a real build, or a week of writing down how the work actually happens.

If you want a second read on it, the free Checklist scores your business in about 15 minutes, and a Discovery Call is a 20-minute conversation, not a pitch.