A practical guide to Intercom collections

A well-organized help center is the difference between happy customers who find their own answers and a support queue overflowing with simple questions. It's the first place people look for help, and when it works, everyone wins. For teams using Intercom, collections are your main tool for bringing order to that chaos. They’re how you turn a jumble of articles into a library that actually makes sense.

But these days, building a great help center isn’t just about making it easy for people to browse. You also need a knowledge base that’s clean and reliable enough to power an AI bot. This guide will walk you through what Intercom collections are, how to get the most out of them, and crucially, how to work around their biggest flaws to build a solid foundation for both your customers and your AI.

What are Intercom collections?

Simply put, Intercom collections are folders for your help articles. They're the main way you can group and organize everything in your Intercom Help Center. Instead of just showing users a massive, unsorted list of guides, you can group related content into logical categories.

Collections have a nested structure, meaning you can have folders inside of folders, up to three levels deep. For example, you could have a main collection called "Getting Started," with sub-collections for "Setting Up Your Account" and "Configuring Your Profile." You could even add another layer for more specific topics under those.

A screenshot showing the folder-like structure of Intercom collections in the Help Center.

The goal here is pretty straightforward: make it easy for customers to find what they need without having to talk to a person. A well-thought-out set of collections guides users to the right answer, which means fewer support tickets for your team to handle.

How to set up and manage your Intercom collections

Getting started with Intercom collections is fairly simple. This isn’t a step-by-step tutorial, but rather a look at the key things you’ll be doing to build and maintain your help center’s structure.

Creating and organizing Intercom collections

You can add a new collection right from your Intercom settings. When you create one, you’ll give it a name, a short description, and an icon to make it stand out. Don't skip the description, it gives users a quick snapshot of what’s inside and helps them find their way around faster.

Once you have some collections and articles ready, organizing them is just a matter of dragging and dropping. Intercom's interface lets you visually reorder articles within a collection or move whole collections around. This makes it easy for anyone on the team, whether they’re technical or not, to manage the content structure as your product changes.

Best practices for structuring your Intercom collections

How you organize your collections has a massive effect on how useful your help center actually is. Intercom points to four common ways to do it, and the best one for you really depends on how your customers think about your product.

1. By product area: If your product has distinct features or modules, this is a solid choice. A project management tool, for instance, might have collections for "Tasks," "Reporting," and "Integrations." It’s clean and predictable.

2. By what users are trying to do: This approach focuses on the user's goal, or their "job-to-be-done." So instead of a collection called "Collaboration Features," you might name it "Working with Your Team." It frames the content around their needs, not your features.

3. By common questions: Grouping your most frequently asked questions into collections like "Billing & Account Management" or "Troubleshooting" is a great way to solve common problems before they become support tickets.

4. By sequential steps: For things that have a specific order, like onboarding or a tricky setup process, organizing articles sequentially just makes sense. You can walk users through the entire process, from start to finish.

Take a minute to think about your users and their journey. Pick the structure that feels the most natural for them, not for you.

Key features and critical limitations of Intercom collections

Intercom collections are a decent starting point for any help center, but you’ll likely start to feel the constraints as your team and knowledge base grow. Knowing what works well and where the platform falls short is important, especially if you’re planning to bring AI into the mix.

What works well with Intercom collections

• Flexible hierarchy: The three-level structure is usually enough for most small and medium-sized businesses. You can map out your product in a good amount of detail without things getting too complicated.

• Simple UI: The interface is clean and easy to use. Your content managers can create, edit, and organize collections without needing to bother a developer for help.

• Multi-help center support: If you manage different brands or products, Intercom lets you use the same article in multiple Help Centers. This is a nice little feature for keeping content consistent across different sites.

Where you'll run into trouble with Intercom collections

And here’s where things start to get tricky.

• One article, one collection: This is the big one. An article can only live in a single collection within the same help center. At first, this sounds like a small detail, but it quickly turns into a major headache. For example, does an article on "How to Reset Your Password" belong in the "Account Management" collection or the "Troubleshooting" collection? Realistically, it belongs in both, but Intercom makes you choose.

• The duplication dilemma: The official workaround for this is to just duplicate the article. But this solution creates a whole new set of problems that are much worse than the original one:

  • A maintenance nightmare: If you need to update that password reset process, you now have to remember to find and edit every single copy of that article. If you miss one, you’re feeding your customers conflicting information.

  • Inconsistent analytics: How do you know if your password article is any good? When you have duplicates scattered everywhere, you can’t get a clear, single view of its performance.

  • Conflicting AI answers: This is the real killer for any automation plans. When you train an AI on a help center filled with duplicate articles, you’re essentially feeding it multiple versions of the truth. The AI gets confused and starts giving inconsistent, unreliable, or flat-out wrong answers to your customers.

• No collection-level restrictions: You can set viewing permissions for individual articles (like making them visible only to logged-in users), but you can’t apply those permissions to an entire collection at once. If you need to lock down a whole section of your help center, you have to do it article by article, which is slow and easy to mess up.

Powering your Intercom help center with AI: The Intercom collections problem

Those limitations we just talked about go from being annoying to a serious problem when you’re trying to build a reliable AI support agent. The main issue is that AI needs a clean, single source of truth to work well, but Intercom's structure almost forces you to duplicate content and create messy knowledge silos.

The problem with a siloed knowledge base from Intercom collections

When your knowledge base is a mess, your AI will be unreliable. It’s that simple. To automate support well, your AI needs to pull from one definitive library of information. The "one article, one collection" rule in Intercom directly undermines this, pushing you into a corner where you have to choose between a confusing help center for your users or a confusing knowledge base for your AI.

Unifying knowledge from Intercom collections without the migration headache

This is where a tool like eesel AI changes the game. Instead of making you move your entire help desk just to get better AI, eesel is designed to plug into the tools you already use and unify your knowledge behind the scenes.

Here’s why that matters: eesel AI can connect to your Intercom Help Center, but it can also pull in knowledge from all your other sources, like Confluence, Google Docs, and even your past support tickets.

This creates a single, unified brain for your AI agent. Your team can keep organizing articles in Intercom in whatever way makes the most sense for customers browsing the help center. Meanwhile, eesel AI is in the background, drawing from a complete, de-duplicated, and much richer knowledge base. You get the best of both worlds: a clear help center for your users and an accurate AI for your support team.

A diagram showing how a unified AI can pull from multiple knowledge sources, overcoming the silo problem of Intercom collections.

Intercom pricing

Intercom collections are a core part of the platform, so you don't pay for them separately. Your access just depends on which Intercom subscription plan you have. The plans are generally based on a per-seat cost plus a fee for each "resolution" handled by Fin, their AI agent.

Plan	Fin AI Agent	Seat Price (Billed Annually)	Key Features
Essential	$0.99/resolution	$29/seat/mo	Fin AI Agent, Messenger, Shared Inbox, Public Help Center
Advanced	$0.99/resolution	$85/seat/mo	Everything in Essential + Workflows, Private Help Center
Expert	$0.99/resolution	$132/seat/mo	Everything in Advanced + SLAs, Multibrand Help Center

It's worth pointing out that a pricing model based on resolutions can get unpredictable and expensive as your support volume grows. You end up paying more as you successfully deflect more tickets, which feels a bit backward.

Build a foundation that scales

So, what's the verdict? Intercom collections are a perfectly fine tool for organizing a help center for your human visitors. They're easy to use and offer enough flexibility for a lot of teams. However, their limitations, especially that "one article, one collection" rule, create real problems for knowledge management and AI automation as you grow.

Duplicating content might seem like an easy fix, but it leads to a tangled mess of maintenance headaches, unreliable analytics, and a confused AI. The smarter approach is to structure your collections for user clarity while using a tool that can unify all your knowledge behind the scenes. That's how you power an AI agent that's actually intelligent and reliable.

Go beyond Intercom collections with eesel AI

You don't have to choose between a well-organized help center and a powerful AI. eesel AI works as a smart layer on top of the tools you already have, including Intercom.

You can keep your help center exactly as it is today. Just connect your Intercom articles, Google Docs, and other knowledge sources to eesel AI to create a single source of truth. The result is a more accurate AI agent that can handle more support tickets without forcing you to switch platforms or completely change how your team works.

Try eesel AI for free or book a demo to see how you can unify your knowledge today.