At most one email a week. Unfollow anytime with one click.
Transcript
The full narration of this video. The current sentence is highlighted while the video plays; click any sentence to jump there.
Show transcript
Introduction
Welcome back to Session 6. In the previous episode we met triggers, actions, and workflows, and built our first practice workflow. Now we open up the nodes: how data flows as items and fields, how to read the input and output panels, how to map fields with drag and drop, and how to test step by step.
Step 1: Data flows as items and fields
Let's start with something familiar: an order form. Rina fills in the first field, her name. Then the product she's ordering, a one-liter palm sugar iced coffee, and the quantity, two. Finally the notes: deliver in the afternoon. When Rina presses submit, this one form submission enters n8n as one package of data. Each label on the form becomes a field name, and each answer becomes its value. Remember these two words, field and value, because we'll use them throughout this session. The easiest way to understand data in n8n is to picture a table. The top row holds the column names: name, product, quantity, and notes. Each row below is one order. In n8n, one row like this is called one item. Rina's row is one item. Three orders means three items. When this data flows to the next node, that node does its job for each item, one by one, automatically. You don't need to build a loop yourself. And each column is called a field. The name field in Rina's item holds Rina, and the product field holds the one-liter palm sugar iced coffee. So the item is the row, and the field is the column. Let's sum up with four key words. First, item: one row of data, for example one order or one form submission. Second, field: one column, like name or product. Third, value: what's inside that field, for example Rina, or the number two. Fourth, flows. Data is passed along like a relay from left to right. The first node's output becomes the second node's input, the second node's output becomes the third node's input, and so on. Items can come from anywhere. From a form, each submission becomes one item, like Rina's order. From Google Sheets, each row read becomes one item. Reading ten rows means ten items flow to the next node. From Gmail, each email becomes one item, with fields like sender, subject, and body. Even an AI answer becomes an item, with fields like summary and category. That's what we'll build in Episode 29. So whatever the source, you always read it the same way. Now let's look at the practice workflow from last episode's homework. The first node is the manual trigger, which runs when we press the test button. The second node is an Edit Fields node we've named Order data. It holds the three sample orders, so we can practice without waiting for real customers. The third node is also Edit Fields, named Build message. Its job is to build a notification sentence from the order data. This practice is exactly what we'll use later for the email to Dimas. After running it, look at the numbers above the lines. One item comes out of the trigger, while three items come out of the Order data node. These numbers are a quick way to check that data is flowing as expected.
Step 2: Reading the input and output panels
Double-click the Build message node. In the real n8n, the node panel opens wide with three parts: input on the left, settings in the middle, and output on the right. In this video we show them one at a time so they're easy to read. The input part shows the data coming into this node from the previous node. There are three items, and the one showing now is the first item, Rina's. Each box in the panel is one field: name, product, quantity, and notes, each with its value. This is the material available to this node. The arrow buttons in the panel let you move to Budi's and Andi's items. The same data can be viewed in three ways. Table looks like a spreadsheet. JSON shows structured text. Schema shows only the list of field names, and it's the nicest view for dragging fields. The output part shows this node's result. For Rina, it's one new field called message, holding the sentence: New order from Rina, two times palm sugar iced coffee one liter. Remember the golden rule: a node's output becomes the next node's input. If the final result is empty or odd, open the nodes from left to right, and look for the first node whose output starts going wrong. To make it easy to remember, think of each node as an employee. The input part answers the question: what material did I receive? The settings part answers: what do I need to do? This is where you choose the operation and map fields. And the output part answers: what did I produce, and is it right? The habit of checking all three will save you from a lot of confusion. So when a workflow doesn't behave as expected, ask these three questions in order at every node: is the material right, is the instruction right, is the result right? Wherever the answer first becomes no, that's where the problem is.
Step 3: Mapping fields: drag and drop
Now the most important part: mapping fields. In the settings of the Build message node, we create a new field called message. Its value is still empty. We want this value to hold a different customer name for each order. So we don't type Rina by hand, because the next order might be from Budi or Andi. Here's how: drag the name field from the input panel, and drop it into the value box. The box changes color, and some text appears between double curly braces: dollar json, dot, name. Don't be scared of this text. It simply means: take the value of the name column from the incoming data. For Rina's item the result is Rina, for Budi's item the result is Budi. Let's finish the sentence. Type plain text like a colon, then drag the quantity field and the product field. Plain text and expressions can be mixed in one box. Under the box, n8n shows a preview of the result for the selected item: New order from Rina, two times palm sugar iced coffee one liter. If the preview is right, your mapping is right. Let's get used to reading expressions. The text dollar json dot name means: take the value of the name column from the incoming data. For Rina, the result is Rina. The pattern is the same for quantity: take the value of the quantity column, and the result is two. And for product: take the value of the product column. So the word after the last dot is always the column name. You don't need to memorize anything, you just need to be able to read it. Sometimes you'll see a longer form, with a node name in brackets. It means take the name column from a specific node, Order data. This form appears on its own when you drag a field from a node further back, and we'll see why it's useful in a moment. This field-mapping skill isn't just for building sentences. In Episode 29, we write a prompt for the AI inside the workflow. The task part is written normally. The data part is filled with dragged fields: name, quantity, product, and notes. Every time a new order comes in, the prompt automatically holds that order's data. Then the format part sets the shape of the answer. So the prompts you practiced in Session 2 are still useful, only now the data is filled in automatically by the workflow.
Step 4: Execute step and executions
After mapping fields, we test. There are two ways. The first way is to test one node with the Execute step button inside the node panel. In some versions it's called Test step. Click Execute step on the Order data node. That node runs, and because it needs a trigger, the trigger before it runs too. A green check mark appears, and its output can be checked right away. Move on to the Build message node. Since its input is already available from the previous test, this node can run on its own. Look at its output: three sentences for three orders. The second way is to click Execute workflow to run everything from the trigger to the last node. That's one execution, and its log goes into the Executions tab. Now an example of a common mistake. The field name was typed by hand and misspelled: nam, not name. The node still runs and still turns green, but look at the result: the customer's name is missing. The lesson: a green check only means the node ran successfully, not that the result is right. Always read the output. The habit of dragging fields instead of typing them prevents typos like this. If there's one sentence to remember from this episode, it's this: a green check is not a guarantee the result is right. A green check only means the node ran without errors. Whether the content of the result is right, you're the one who checks it in the output. Make it a habit to open the output, at least for the first item. Let's line them up. Execute step runs one node, plus earlier nodes if their data isn't there yet. Use it every time you finish setting up a node. Execute workflow runs the whole flow from the trigger. Use it to test the complete flow before using it for real. And an active workflow runs on its own every time the trigger fires, without you pressing any button. We'll only turn this on in Episode 30, after all tests pass. One more note. Test results from Execute workflow are usually also logged in the Executions tab, marked as tests, depending on your execution-saving settings. That way you can reopen yesterday's test results to compare.
Step 5: Arranging blocks: naming and inserting nodes
One small habit that helps a lot: give nodes clear names. When added, nodes are automatically named things like Edit Fields and Edit Fields one. A week from now, you'll surely forget which is which. Right-click a node and choose rename, or press F2, then type a name that explains its task: Order data, and Build message. These names are also used by expressions that refer to a specific node. What if you need a new step in the middle? Hover over the line between two nodes, then click the plus sign that appears. For example, let's insert an Edit Fields node called Add time, to record the order time. Let's run it again. All nodes are green, but open the Build message output: the name, quantity, and product are empty. Only the sentence skeleton is left. Why? By default, the Edit Fields node only passes on the fields it creates itself. The Add time node only outputs the time field, so name, product, and quantity aren't passed along. Fix number one: in the Add time node, turn on the option to include other input fields. That way the old fields are passed along together with the time field. Fix number two: take it directly from the Order data node. In the input panel, drag the name field from that node's section, and the expression automatically names the node. It means: take the name column from the Order data node. Run it once more. The sentence is complete again, now with the order time. This is why node names should be clear: an expression that says Order data is much easier to read than Edit Fields one. So which fix should you choose? For beginners, fix number two is safer, because it doesn't depend on the settings of a node in the middle. When you drag a field from an earlier node, n8n already writes it correctly. Let's collect the good habits. First, name each node after its task. Second, test each node as soon as it's set up, don't wait until everything is finished. Third, drag fields from the input panel, don't type them. Fourth, read the output, don't settle for just a green check. Fifth, use the pin feature. A node's output can be pinned as sample data, so when testing the next node you don't need to fill in the form or re-trigger from the start. Sixth, save the workflow as often as possible. Seventh, when practicing, use made-up data like Rina and Budi, not real customer data, and certainly not real phone numbers and addresses.
Step 6: Exercise: apply it to your own data
The pattern we just learned works for any job. Take Mr. Hendra at Cerdas Tutoring. After every practice test, he sends each student their score. His data is a table with the same pattern: name, class, score, and notes. Raka in class nine A scored eighty-six, up ten points. Three students means three items. With a workflow, Mr. Hendra only has to build one sentence, and n8n creates it for every student. The workflow is exactly like the Sunset Coffee practice: a manual trigger, a Score data node, and a Build update node. In the Build update node, drag the name, class, and score fields into the message box, then add plain text around them. The expressions read: take the value of the name column, the class column, and the score column. Notice that plain text like the word Hi and Notes is typed directly, while name, class, score, and notes are dragged from the input panel. You'll use combinations like this often. Run it, and look at the output: three personal updates, each with its own name, score, and notes. In Session 8, updates like these can be sent straight through WhatsApp or Telegram. Now it's your turn. First, finish the Sunset Coffee Practice workflow until the Build message node runs for three orders. Second, build the sentence by dragging the name, quantity, and product fields, then make sure the preview is right. Third, make a version for your own work, whether it's student, client, or product data, with at least three items. Fourth, save the workflow, and note one thing that still confuses you. Usually that confusion gets cleared up when we practice with real data in the next episode.
Done!
Let's sum up. Data in n8n flows as items and fields. The node panel shows input, settings, and output. Drag fields to create expressions, and read them as take the value of a certain column. Then test each node, and always read the output. In the next episode we start building a real workflow: the Sunset Coffee order form becomes the trigger, and we watch the first order's data appear in the panel. See you in Episode 28.
Practice exercises
Try these on your own after watching to lock in what you learned.
Add a new field called greeting in the Build message node, containing 'Hi' followed by the customer's name, dragged from the name field.
Insert an Edit Fields node in the middle of the flow, run it, and see for yourself what happens when the option to include other input fields is switched off and on.
Build a second practice workflow with data from your own work (students, clients, or products), then build one notification sentence from three fields.
Ask the class material
Type a question or keyword; we find the part of this episode that explains it and jump the video there.
My notes
Notes are saved with the current video time and are only visible to you.
Questions & answers
Stuck on a step? Ask here; the instructor and other learners can answer.
No questions yet.
Your instructor
Cerdas Inovasi Digital Team
Software developers, BIM & 3D engineers, and AI practitioners
53 video classes1,278 episodes
We build software, BIM models, games and AI automations for clients every day, and turn that hands-on work into step-by-step classes: every step is shown on screen and comes with a PDF guide you can keep.
