A practical guide to using an Intercom workflow to route first messages to the right team

Manually routing customer messages is one of those tasks that seems small but ends up eating a huge chunk of your day. It’s slow, a perfect recipe for human error, and frankly, a bit of a drag on your support team's morale. Every minute an agent spends playing traffic cop for tickets is a minute a customer is left waiting for an actual answer.

Intercom’s Workflows feature is a built-in way to automate this, especially for those all-important first messages that can make or break a customer's experience.

This guide will walk you through how to set it up and what it can do. But we’ll also be honest about where it hits its limits, especially for more complex support teams, and show you a more flexible alternative for when you’re ready to take your automation up a notch.

What exactly is an Intercom workflow?

Intercom Workflows is the platform’s tool for building automations without needing to write any code. Think of it as a way to draw a flowchart for your customer conversations. You pick a trigger, like "a customer sends a new message," and then line up a series of actions, like "assign this message to the sales team."

A visual of Intercom's no-code workflow builder, which allows users to create automations for routing messages.

Its main job is to get repetitive tasks off your team's to-do list. This could be as simple as routing and tagging conversations, or a bit more involved, like letting Intercom’s AI agent, Fin, take the first crack at an answer.

And because it’s baked right into Intercom, it works across all the ways customers contact you, live chat, email, you name it, which helps keep the experience consistent.

How to set up an Intercom workflow to route first messages to the right team

Getting a basic routing workflow up and running is pretty straightforward. We won’t go through every single click, but here’s a quick look at the main pieces you’ll be working with to build your "Intercom workflow to route first messages to the right team".

Start with a trigger

Every workflow needs a starting gun. For this job, the best trigger is usually "When a customer sends their first message." This makes sure your automation springs into action the second a new conversation starts.

When you're setting this up, you can also get specific about the audience. For instance, you could have the workflow only apply to conversations from your website's live chat, or you could make it apply to everyone, including folks emailing your support address. You can also slice it by customer type, maybe applying one set of rules for paying customers and another for anonymous visitors.

Use branches to create routing paths

After the trigger fires, the real decision-making happens in the "Branches." This is where you build the "if this, then that" logic that sends conversations down different paths.

Here are a few common ways people use conditions:

• Customer Data: You can use info you already have about your customers to route them. If a customer is on your "Enterprise" plan, you can create a branch that sends them straight to a dedicated support team. No waiting in the general queue.

• Detected Language: Intercom is smart enough to detect the language of an incoming message. You can use this to automatically send a question in French to your French-speaking agents. Simple, but it makes a huge difference to the customer.

• Office Hours: What happens if a message comes in at 2 a.m.? You can set up a condition based on your office hours. This path could send the ticket to an international team in another time zone or just fire back an automated reply letting them know you'll get back to them in the morning.

Assign conversations to the right team

The last piece of the puzzle is the action. At the end of each branch, you tell the workflow what to do. In our case, you'll use the "Assign" action. This lets you point the conversation to a specific team inbox, like "Sales," "Tech Support," or "Billing."

For this to work, you’ll need to have your teams set up in Intercom first. Once that's done, you just pick the right team from a dropdown menu for each path you've built. The conversation will land in the right place, ready for the right person to grab it.

Going beyond basic routing

Once you’ve got basic routing down, you can use workflows to handle a few other clever tasks.

Let Fin AI have the first word

A pretty neat trick is to put the "Let Fin Answer" action right at the top of your workflow. When a new message arrives, Fin, Intercom's AI agent, immediately scans your Intercom Help Center to see if it can find an answer.

An example of Intercom's Fin AI Agent answering a customer query via email, showcasing the 'Let Fin Answer' feature in action.

If Fin nails it and solves the problem, the conversation gets closed without ever bugging your team. This is a great way to deflect all those common, repetitive questions and free up your agents for the trickier stuff. And if Fin is stumped? No problem. The workflow just continues down the path you built, routing the customer to the best-equipped human for the job.

Automate inbox cleanup

Workflows can also be your behind-the-scenes janitor, handling the kind of admin tasks that nobody enjoys. This helps keep your inbox tidy and makes sure small but important details don't get missed.

A couple of examples:

• Automatic Tagging: You can create rules to automatically add tags based on keywords. If a message contains the word "refund," the workflow can tag it as "Billing Inquiry," which makes reporting way easier down the line.

• Applying SLAs: If a customer is on a premium plan, a workflow can automatically apply the right Service Level Agreement (SLA), making sure your high-priority customers get the speedy responses they're paying for.

• Closing Inactive Conversations: You know when you ask a customer for more info and then... crickets? A workflow can send a polite nudge after a few days and then automatically close the conversation if you still don't hear back.

Where Intercom workflows start to creak

Intercom Workflows are a great place to start, but as your support operation gets bigger and more complex, you might start to bump up against some of its limitations.

Why the setup can be rigid

While the visual builder feels simple at first, a workflow with dozens of routing rules can start to look like a tangled plate of spaghetti. According to Intercom’s own help guides, it’s easy for conflicting rules to cause a workflow to break, leaving conversations stuck in limbo. Trying to figure out what went wrong can be a real headache.

This is a common growing pain, and it’s why more modern tools take a different tack. With eesel AI, for example, the goal is to get you started in minutes. Instead of making you build every single rule from scratch, it learns from your team's past support tickets to figure out routing patterns on its own, which cuts down the setup work quite a bit.

Knowledge limitations

Intercom's AI, Fin, does its best work when it's reading from content inside Intercom, like your Help Center articles. This is a problem for most companies, where knowledge is spread out all over the place. If your best troubleshooting guides live in Confluence, your internal processes are in Google Docs, and your FAQs are in Notion, Fin can't easily access any of it.

This is exactly the kind of problem eesel AI was designed to handle. It doesn't just plug into your helpdesk; it connects to all the different places your team stores information. It reads your help articles, internal wikis, and even past ticket resolutions to get the full picture, which leads to much more accurate answers without you having to migrate all your content.

A diagram comparing the limited knowledge sources of a built-in AI like Intercom's versus an agnostic AI that connects to multiple sources.

No risk-free way to test workflows

When you build a new workflow in Intercom, it's tough to know how it will actually behave with live customers. You can test little pieces of it, but there isn't a good way to simulate how the whole thing will perform against thousands of your real-world conversations. You kind of have to flip the switch and hope for the best.

In contrast, eesel AI has a simulation mode that lets you safely test everything beforehand. You can run the AI on thousands of your past tickets and see exactly how it would have handled each one. This gives you a solid prediction of your deflection rate and helps you spot gaps in your knowledge base before a single customer interacts with it. It’s about replacing guesswork with confidence.

A quick look at Intercom's pricing

Alright, let's talk about the money part. Understanding Intercom's pricing can be a bit confusing because it’s made up of a few different pieces: a platform subscription, a fee for each agent, and charges for how much you use its AI features.

Here’s how it usually breaks down:

• Platform Plans (Essential, Advanced, Expert): You pick a plan that gives you a set of features, and you pay a monthly fee for every agent on your team. More agents means a higher base cost.

• Fin AI Agent: This is where it gets tricky. Intercom charges $0.99 for every conversation that Fin successfully resolves on its own. This means your costs can be unpredictable and might jump during busy times.

• Add-ons: Many of the more advanced features, like proactive support or the agent-assist tool Copilot, will cost you extra each month.

The main thing to know is that the pay-per-resolution model can make budgeting tough. As your AI gets smarter and solves more tickets, your bill goes up. You're effectively charged more for being successful.

A more flexible alternative: eesel AI

If the limits of Intercom Workflows and its pricing feel a little restrictive, eesel AI offers a different path. It’s built to solve the exact challenges that growing support teams run into.

Here’s what makes it different:

1. It works with your existing helpdesk. No need to move everything over. eesel AI plays nice with the tools you already use, including Intercom, Zendesk, Freshdesk, and many others.

2. It brings all your knowledge together. eesel AI connects to all your information sources, help articles, Google Docs, Confluence, past tickets, to give your AI the full context it needs to give genuinely helpful answers.

3. You get total control and safe testing. With its simulation mode, you can test your setup on thousands of past tickets, get reliable performance numbers, and roll out automation with confidence, starting small and scaling up when you're ready.

4. The pricing is transparent and predictable. eesel AI plans are based on a flat monthly fee for a set volume of interactions, with no per-resolution charges. Your bill doesn't shoot up just because you had a busy month, making it much easier to budget as you grow.

Automate smarter, not harder

Setting up an "Intercom workflow to route first messages to the right team" is a great first step toward a more efficient support system. For any team still doing things manually, it's a no-brainer.

But as your team gets bigger and customer questions get more complex, you'll probably feel the constraints of a closed system, scattered knowledge, and unpredictable pricing. Those are the kinds of challenges that can keep you from providing the fast, accurate support that customers have come to expect.

When you're ready to move past those barriers, a tool like eesel AI is the next logical step. It’s a more integrated and cost-effective AI solution that works with the tools you already have, giving you the freedom to automate smarter.

Ready to see how easy AI-powered support can be? Try eesel AI for free and see it work on your own tickets in just a few minutes.