A complete guide to the Intercom Fin Topics Explorer in 2025

Ever feel like you're just guessing what your customers really need? You know the feeling. You're trying to get a handle on the top support issues, but manual tagging is a mess, it's slow, inconsistent, and usually the first thing that gets dropped when things get busy. Without a clear map of what’s driving your ticket volume, you’re basically flying blind.

Intercom’s solution for this is the Fin Topics Explorer, an AI tool that’s supposed to automatically make sense of all your support conversations. The promise is that it will show you what’s actually happening in your inbox, with zero manual work.

In this guide, we'll walk through what the Fin Topics Explorer is, how it works, and where it’s genuinely useful. But we'll also get real about its key limitations and explore how modern alternatives give you the power and flexibility to not just analyze your support data, but act on it with confidence.

What is the Fin Topics Explorer?

At its core, the Fin Topics Explorer is an AI-powered analytics feature that comes bundled with Intercom’s Fin AI Agent. Its job is to automatically sort your support conversations into topics (like "Billing Questions") and even more specific subtopics (like "Refund Request"). The best part? No one on your team has to lift a finger to tag anything.

Think of it as an automated dashboard giving you a 30,000-foot view of your support queue. Instead of relying on gut feelings or clunky spreadsheets, the explorer gives you a live, data-driven look at what your customers are asking for most.

A look at the main dashboard of the Intercom Fin Topics Explorer, which provides an overview of support conversation analytics.

The goal is to help you see trends as they emerge, track how you’re handling certain problems, and find areas where your support team could be doing better. It’s about taking that messy flood of daily conversations and turning it into a clear picture of what's happening, right inside your Intercom ecosystem.

How the Fin Topics Explorer works

So, how does it actually turn a mountain of customer chats into a neat, organized dashboard? It’s all powered by machine learning, and it happens in a few steps.

AI-powered topic and subtopic discovery

First, the system dives into your past conversations, up to 90 days' worth, to learn the unique ways your customers ask for help. It uses machine learning to find and group similar questions.

What's interesting is that it figures out the specific subtopics first. For example, it might spot a bunch of questions like "How do I reset my password?" and "I'm locked out of my account." From there, it groups these related subtopics into a bigger topic, like "Account Access." The AI comes up with clear titles for these categories on its own. This isn't just about matching keywords you've set up; it's about understanding the actual meaning behind the questions.

Conversation assignment and key criteria

Once the topics are identified, the system starts assigning them to conversations in two ways:

1. Backfilling: It combs through the last 90 days of your conversations and applies the new topics to them retroactively.

2. Inference: Every day, it looks at newly closed tickets and sorts them into the right topic buckets.

For a conversation to even be considered, it has to meet a few rules. It can’t be spam, it needs to involve at least two people (a customer and an agent), and there must be at least 15 similar conversations to create a subtopic. If your ticket volume is low or the questions are all over the place, you might not see any topics show up at all.

Visualizations and performance metrics

All these insights get served up in a dashboard with two main views:

• Treemap: This is a block of colored boxes. The size of a box tells you the conversation volume for that topic, and its color changes based on a metric you choose (like customer satisfaction). It's a quick way to see which high-volume topics are causing the most trouble.

• Ridge Line Charts: These charts let you see how the volume and performance for each topic have shifted over time, making it easy to spot sudden spikes or downward trends.

You can filter this data by metrics like CX Score, how often Fin was involved, Fin's resolution rate, and median handle time.

Having a dedicated analytics view is helpful, but these insights are stuck inside a single platform. You're only seeing one piece of the puzzle. This is where a tool like eesel AI works differently by pulling knowledge from everywhere your team works, including outside sources like Confluence or Google Docs. This way, your analysis isn’t siloed and actually reflects all of your company's knowledge, not just what’s sitting in your helpdesk.

Key use cases for the Fin Topics Explorer

Knowing what the Fin Topics Explorer does is one thing, but how does it actually help day-to-day? Here are a few ways teams put this kind of analysis to work.

Spotting trends and catching problems early

By keeping an eye on the topic dashboard, you can catch problems before they blow up. Imagine you see a sudden spike in conversations under the "Account Locked" topic, and the customer satisfaction scores for those tickets are tanking. That’s a massive red flag that a recent product update might have shipped with a bug. You can jump on it and alert the engineering team immediately, instead of waiting for a wave of angry customers.

Identifying areas for optimization and training

Topic analysis is great for figuring out where to focus your improvement efforts. Let's say you notice that a high-volume topic like "How to set up X feature" has a low AI resolution rate and your agents are spending a long time on each ticket. That’s a clear sign that you’ve got a knowledge gap.

You can use that insight to take direct action. Maybe it’s time to write a new, super-detailed help center article. Or perhaps you need to build a better automated workflow for your AI agent. It could even be a signal that your team needs a training session on how to handle those questions better.

Filtering reports for deeper insights

The topics the AI finds can also be used as filters across other Intercom reports, letting you dig deeper into your data. For example, you could filter a team performance report by the "Refund Request" subtopic. This would let you see how different agents are handling those tricky conversations and maybe even identify a top performer whose methods could be taught to the rest of the team.

Acting on these insights is where the real value is. But instead of just flagging a knowledge gap, a platform like eesel AI goes a step further by automatically generating a draft help article from a successfully resolved ticket. It also has a powerful simulation mode where you can test any improvements on thousands of your past tickets. This lets you see the direct impact on resolution rates before you roll out a single change, turning a good idea into a proven win.

Limitations of the Fin Topics Explorer

While the Fin Topics Explorer is a decent starting point, it has some pretty big limitations that can get in the way once you start trying to seriously automate your support.

No customization or direct control

This is the big one, and it's mentioned right in Intercom's own help docs: "It’s currently not possible to modify or edit the AI topics/subtopics." That's a direct quote. It means you're stuck with whatever the AI decides, even if its categories don't quite match how your business works. You can't merge two similar topics, split a topic that’s too broad, or fix a conversation that was obviously put in the wrong bucket. You get what the "black box" gives you, and that's it.

This lack of control can be really frustrating. In contrast, a tool like eesel AI is built around a fully customizable workflow engine. It puts you in the driver's seat, letting you create automation rules to decide exactly which tickets the AI should handle. You can start small with easy, high-confidence topics and have the AI pass everything else to a human, then slowly expand its duties as you see how it performs.

The Fin Topics Explorer's closed knowledge ecosystem

The Fin Topics Explorer only looks at what's happening inside Intercom. That's a huge blind spot. For most companies, the best and most up-to-date information isn't in old support tickets, it's scattered across internal wikis, product docs, and team chats. If your team lives in Confluence, Google Docs, or Slack, the AI can't see any of it. This means your analytics will always be incomplete, and any automation you try to build will be based on just a tiny fraction of your company's knowledge.

eesel AI is designed to unify your knowledge, not trap it. It connects to over 100 sources, helpdesks, wikis, chat tools, you name it, giving your AI a complete picture of your business right from the start.

The Fin Topics Explorer's 'black box' deployment

With the Fin Topics Explorer, you only see the analytics after the fact. There’s no good way to test or predict how a change, like adding a new help article or tweaking a workflow, will affect performance before it’s live with your customers. You’re essentially making changes in the dark and just hoping for the best.

eesel AI takes the guesswork out of the equation with its simulation mode. You can test your AI setup on thousands of your own historical tickets to get a data-backed forecast of its resolution rate. You can see exactly how it would have answered real customer questions, letting you fine-tune its behavior and roll out changes with total confidence.

Feature	Fin Topics Explorer	eesel AI
Topic Customization	None, topics are fixed	Full control via selective automation rules
Knowledge Sources	Intercom conversations only	Help desk, Confluence, Google Docs, Slack, & 100+ more
Pre-Launch Testing	Post-launch analytics	Powerful simulation on historical tickets
Setup & Integration	Locked into Intercom suite	Radically self-serve, connects to existing tools in minutes
Custom Actions	Limited to Intercom features	Trigger API lookups, update ticket fields, and more

Intercom Fin pricing explained

The Fin Topics Explorer isn't something you can buy on its own. It’s a feature that’s part of the Fin AI Agent, which is an add-on you can purchase on top of Intercom's regular customer service plans.

Fin’s pricing is based on performance: it starts at $0.99 per resolution. A "resolution" is counted every time the AI successfully answers a customer's question without needing to escalate to a human agent.

On the surface, paying per resolution sounds fair. But it has a tricky downside: your costs are totally unpredictable and grow as your ticket volume does. If you have a busy month or the AI gets better at its job, your bill goes up. You’re essentially paying more for being successful, which can make budgeting a real headache and might even make you hesitant to automate more of your support.

The eesel AI alternative: Transparent and predictable

We do things differently at eesel AI. Our pricing is straightforward and predictable. Our plans are based on a set number of "AI interactions" per month (which could be an AI reply or a workflow action), and we have zero per-resolution fees. You will always know exactly what your bill will be, no matter how many tickets your AI resolves.

This transparency gives you the freedom to automate as much as you want without dreading a surprise bill. Plus, you can start with a flexible month-to-month plan and cancel anytime. Many competitors try to lock you into long annual contracts from day one, but we believe you should be able to prove the value without a risky commitment.

Fin Topics Explorer: Move beyond analytics with a unified AI platform

Look, understanding why customers are reaching out is a great first step. Tools like the Fin Topics Explorer can definitely give you a starting point. But just looking at data isn't enough. It's held back by a lack of control, a closed system, and an approach that forces you to test on live customers.

To really improve your support, you need more than just a dashboard. You need a platform that gives you complete control over automation, connects to all of your team's knowledge, and lets you test everything without risk before it ever interacts with a customer.

That’s what eesel AI is all about. It's built for teams that are ready to move from passively looking at their support data to actively and confidently automating it.

Ready to gain total control over your support automation? Start your free eesel AI trial or book a demo to see our powerful simulation mode in action.