Hello and welcome to video number four of the Advanced Course for n8n. In this video, we'll be covering sub-workflows, the secret to making your n8n operations a lot more scalable, but also increasing the speed of workflow building.

So, the first topic we need to cover is, you know, what is a sub-workflow? The Execute Workflow node allows you to call one workflow from another, allowing you to either execute multiple workflows back to back or extract a part of a workflow into its own dedicated workflow.

For example, this workflow has a group of nodes that are used to look up a user in a database. Everything here that is covered by the node, we could replace this with a sub-workflow.

From a technical point of view, it executes the chosen n8n workflow by taking its input data and outputting it from the "When Called by Another Workflow" trigger.

This workflow will then execute item by node by node, item by item, and the items from the final node of the workflow are returned to the original workflow.

So here, in this diagram, the output data from Edit Fields would go into the Execute Workflow.

This would then be used as input data for the Execute Workflow trigger, and when this workflow is done executing, it'll take the items from the last node of the workflow and use it as output data for the Execute Workflow node, and then it'll just continue in the original workflow.

So the sub-workflow pretty much allows us to extract a set of nodes from a given workflow and make it into its own workflow. But why is that useful?

There are some sets of actions that you will often have to do in multiple different workflows, for example, enriching information about a person, looking someone up in a database, looking up a certain order number in a table, and creating a sub-workflow allows you to abstract a more complex task, allowing everyone to use the right method simply by calling the subworkflow.

This also makes maintenance a lot easier. If you need to update the enrichment process or if you need to add a new table in the lookup, it can be done once and the updated version will be used across all workflows that call it.

There are two important things to note when using subworkflows. First, be sure to use a standard format for key names or embedded JSON object structures.

If the input data has a key 'email' with an uppercase 'E', but the subworkflow reads a key 'email' with a lowercase 'e', the subworkflow will fail.

Here, we can see the output data has a capital or uppercase 'E' and the subworkflow is checking for a lowercase 'e'.

The output data of the execute workflow node would also be the output data of the last node of the workflow, so be mindful of which node is the last one in your workflow, what the different values of the keys are, and if you need to join the input data or not.

So, to illustrate how a subworkflow works, um, we're going to use this, um, little workflow that I have here. So I'm just going to test the workflow and on myself a webhook.

What this workflow does is it, uh, extracts the information from the webhook. If it has an ID, it tries to cross-reference it with a Google Sheet and, um, gets any potential matches.

If it doesn't have an ID but has an email, then it does the same operation in the same Google Sheet, this time using the email as the key.

And if no ID and no email are in the webhook, this means it will return an an error.

And so we use the Stop and Error node. Um, then we just send some messages. The whole point of a sub workflow is this lookup would be useful in other places, in other workflows, than only this one.

So what we could do is extract this and replace it by a subworkflow that we would then call from the original workflow.

So here, I'm going to take all of these nodes. I'm going to cut them out and replace them by an execute workflow.

For now, we're just going to leave this empty because we need to create the workflow to get the ID.

Save this workflow, create a new workflow called lookup. Here, I will paste all of the nodes from earlier and launch the workflow with the execute workflow trigger.

One more thing we need to be careful about is the final node. So what I like to do, just to make it a little bit easier, is drag everything here to have a very clear final node.

And then I can get the workflow ID from the URL and heading back to the original workflow, I can now insert this and run the workflow as we did previously.

As we see, it had exactly the same output. It did the whole lookup and returned the right contact.

So this is very useful when we have parts of workflows that we might want to execute multiple times.

A very important thing to note is before using an execute workflow, make sure to standardize the fields, bring them all back top level, just so you don't have any trouble using the subworkflow.

You can also in the subworkflow add some documentation with a note, for example, saying, "Please give me an ID or email field top level." And that way, anyone who wants to use this subworkflow will be able to very easily.

Thanks for listening to the fourth video of the advanced course, where we covered subworkflows, when to use them, and some tips and tricks.

In the next video, we'll look into an advanced error workflow.

n8n Advanced Course (4/8) - Subworkflows

n8n Advanced Course (4/8) - Subworkflows

Scripts