Your First n8n tutorial: Build a Useful Workflow Tonight

اقرا بالعربي

A practical n8n tutorial for building a request form in one evening, cleaning its fields, saving each submission to a table, testing common errors, and planning a safer next step before connecting WhatsApp or real customer data.

Agents Automation
A Cairo shop owner connects a simple gold workflow beside a phone and order slip.

A clothes-shop owner in Cairo may spend the evening copying order details from messages into a notebook, then miss someone who needs follow-up. This n8n tutorial helps you build a small first project: a form collects requests, a workflow tidies the information, and a table saves each submission instead of leaving it in a chat thread.

We will not start with a bot that replies to every customer or connect a live WhatsApp number. The goal is a smaller, testable step: a form that captures information and stores it in an n8n Data Table. Once you understand how data moves between nodes, you can add a condition, a human review step, or an external service more safely.

What will this n8n tutorial build?

A workflow is a route: an event starts it, nodes shape the data, and a final node stores or acts on the result. Here, someone submits a form. Edit Fields gives the incoming information consistent names, and Data Table saves it as a row. A shop owner can then review requests in one place instead of searching through separate conversations.

We will use n8n Form Trigger to create the form page and start the workflow when someone submits it. Then we will use Edit Fields and Data Table. This is not a direct WhatsApp integration: the customer fills in a form. Start with only the details you need, such as a customer name, contact number, request type, and a short description. A useful n8n tutorial keeps this first data path separate from later integrations.

Plan the fields before opening the editor

Write down the information you will use to follow up. For a clothes shop, the form could ask for:

  • Customer name, as required text.
  • Phone number, required if the team will contact the customer.
  • Request type, with choices such as product question, purchase request, or exchange.
  • Details, a short note about the size, colour, or question.

Use short internal names such as customer_name, phone, request_type, and details, even if the customer-facing labels are in Arabic. Those names make it easier to map the data in later steps. Use invented test details, not real customer phone numbers.

Set a clear success test: after submitting a test form, one row appears with the right values in the right columns. This gives you a simple way to know whether the first version works.

Build the workflow step by step

  1. Create a new workflow and name it for its purpose, such as “Shop Requests — Test”. A clear name helps when you have more than one experiment.
  2. Add n8n Form Trigger as the first step. Give the form a plain title and a brief description, add the fields you planned, and mark essential fields as required. Do not promise a reply time your team may not meet.
  3. While building, use the Test URL or form preview. Submit a sample such as “Salma”, an invented phone number, “product question”, and “Is size M available?”. n8n shows the incoming data in the editor during testing, so you can inspect the field names and values.
  4. Add Edit Fields after the form. Create organized fields such as customer_name, phone, request_type, and details, then map each one to the matching form value. A consistent record will be easier to store and connect to another service later.
  5. Add Data Table after Edit Fields. Create a table for shop requests with matching columns, choose the action to insert a row, and map the outgoing fields to those columns. Each node has one job, which makes errors easier to find.
  6. Run the workflow and submit the test form. Inspect the last node’s output: one row should appear, with each value in the right column. Submit another sample to check that a new row is added rather than replacing the first.
Test request: customer name “Salma”, an invented contact number, request type “product question”, and details “Is size M available?”. After submission, those values should appear together in one row. Do not use real customer data just to test the route.

Test the flow before relying on it

A workflow that ran once is not automatically ready for daily use. Try a normal request, one with an optional field blank, and one with a longer description. Inspect the data that arrived. Do not assume the label on the form is the same as the internal field name. Check each node’s execution data to see whether a problem began in the form, the field mapping, or the table.

When the basic route works, add an If node for one clear condition, such as “request_type is exchange”. Route that case for a person to review rather than letting the workflow decide for the customer. Avoid vague rules like “this is important” until you can define them with a testable example.

That is a small example of agentic automation: a defined sequence of steps and responsibilities, not an agent free to make every decision. Keep a person in the loop for sensitive cases.

Use the Test URL while building. When the result and access settings look right, publish the workflow and use its Production URL. The production form runs the workflow after publishing; review those executions in n8n’s execution history. Keep a test copy separate from any workflow that handles real requests.

If you later want an email alert or a Google Sheets record, add that integration after the form-to-table route is stable. It may require credentials and permissions for the external account. Never put account credentials in a public form, and test with sample data before using real shop information.

A Cairo shop example

Imagine a small shop that takes orders through Instagram and WhatsApp. The owner wants to see who asked about a size, who wants to buy, and who has an issue with an earlier order. A simple form gathers requests and records each one in a table. In the morning, the owner sorts by request type and handles the cases that need a person. The workflow has removed a manual copying step; it has not claimed to reply automatically.

After testing with sample data, the owner can decide whether colour is needed, whether an address is needed before confirming availability, and whether an exchange must go to a staff member. Those answers determine the next change. If the shop later needs a WhatsApp AI agent for small business, it already has a clear view of the inputs, expected result, and cases that need a human.

For a later messaging project, read the existing guide to a WhatsApp AI agent for a small shop. It explains response boundaries and when a conversation should move to a person. It is a useful next topic, but it is not required for tonight’s test.

Grow it without making it fragile

After the first evening, ask which repeated task costs you time every week. You might add a notification for exchange requests, format a record for a follow-up team, or connect a data source you already use. Do not add several channels and rules at once. Change one thing, test it with a few sample records, and keep the working version before making another change.

Write a short note explaining what starts the workflow, which fields are required, where the result is stored, and when a person must review it. Ask someone else on the team to submit a test request without your help. If they cannot tell where to inspect the result, simplify the workflow before connecting it to daily work.

For practical learning, explore Etalem AI courses and choose a topic that matches the stage you have reached. Etalem AI is an Arabic platform where working experts teach practical AI; its community lets members share projects, ask questions, and talk with one another and the experts. Use what you learn to understand your options, then connect your own systems only after testing the rules you need.

FAQ

Do I need to know how to code?

Not for this first route. You can choose nodes, connect them, and map fields in the editor. Understanding the incoming data and testing each step matter more than writing code for this project.

Does this connect n8n directly to WhatsApp?

No. This version uses an n8n form and saves submissions in a Data Table. Connecting a WhatsApp channel requires configuring the relevant integration and permissions, then testing incoming and outgoing messages separately. Do not claim that replies are automated until the real route has been tested.

What should I check if no row appears?

Inspect the Form Trigger output first. Then check Edit Fields for correct names and values, followed by the selected table and insert-row action. Review one node at a time instead of rebuilding the whole workflow.

Start with a small version tonight

Choose one request that repeats, create a form with four fields, and test it with invented data until a clean row appears in the table. When you are ready for the next step, browse the n8n courses here. Bring your question or project to the Etalem AI community. Keep a person responsible for reviewing the result, and expand the workflow only after the first version proves useful.

Read next