A complete overview of the Intercom platform in 2025

Kenneth Pangan
Written by

Kenneth Pangan

Stanley Nicholas
Reviewed by

Stanley Nicholas

Last edited October 2, 2025

Expert Verified

If you’ve spent any time in customer support or SaaS, you’ve definitely run into Intercom. It’s one of the biggest names on the block, promising an all-in-one platform that juggles everything from support and sales to marketing. But for a lot of a teams, using it day-to-day brings up a few common headaches: costs that are high and unpredictable, a setup process that can drag on, and AI features that don’t quite give you the control you really need.

These issues are often what push teams to start looking around for other options, ones that offer powerful automation without the pain. This guide will walk you through a detailed, no-fluff look at Intercom’s main features, its AI agent "Fin," its famously complicated pricing, and the limitations you should know about before signing on the dotted line.

What is Intercom?

At its heart, Intercom is a customer communications platform. It’s all about helping businesses chat with their customers using a bunch of connected tools. The main idea is to manage conversations from all over the place, all in one spot.

Its main pieces are:

  • The Messenger: You know that little chat widget that pops up on websites and in apps? That’s it. It’s for real-time conversations.

  • Shared Inbox: This is the command center where support, sales, and marketing teams can see and reply to customer messages from email, chat, and social media.

  • Help Center: A tool that lets you build out a knowledge base so customers can find answers themselves.

  • Fin AI Agent: This is Intercom’s own AI, built to handle questions and solve problems automatically.

A screenshot of the Intercom Messenger, showcasing its modern design.
A screenshot of the Intercom Messenger, showcasing its modern design.

It’s a popular tool for companies big and small, but as we’re about to see, an all-in-one tool isn’t always the best-in-everything tool.

A deep dive into Intercom’s core features

The Intercom platform is built around a few key parts that are supposed to work together. They’re powerful, for sure, but it’s worth understanding how they actually function and where you might find some gaps.

The messenger and inbox

The Messenger is what your customers see; the shared Inbox is where your team lives. The inbox pulls in messages from your site, app, email, and social media into one stream. It has a basic ticketing system, lets you set up different inboxes for different teams, and has rules for routing conversations.

A view of the Intercom ticketing system with customer details shown in the sidebar.
A view of the Intercom ticketing system with customer details shown in the sidebar.

Having one inbox for everything sounds great in theory. In practice, it can quickly become a bottleneck. Without some serious automation, teams often feel like they’re drowning in manual tasks, tagging conversations, routing tickets, and sorting through issues. This is where an AI layer like eesel AI can make a world of difference. Instead of making you switch platforms, it plugs right into your existing Intercom inbox to handle the repetitive stuff, freeing your team up to focus on the conversations that actually need a human touch.

Help Center and knowledge management

Intercom’s Help Center lets you build and host a knowledge base full of articles to help customers help themselves. You can make it public for everyone or keep it private for your team, and it supports different languages.

An image of the Intercom knowledge base interface.
An image of the Intercom knowledge base interface.

The catch? Your knowledge base is only as good as the content you painstakingly write and maintain. It’s a manual, time-consuming job. Teams have to guess which articles are needed most and then spend hours keeping them up to date. This feels a bit old-school compared to modern AI tools like eesel AI, which can automatically sift through your past support tickets and draft new articles based on what actually solves customer problems. It can also pull knowledge from all the other places your team already keeps it, like Google Docs or Confluence, creating a single source of truth for your AI without you having to move a single document.

Proactive support and outbound messaging

Intercom also gives you tools for reaching out to customers first, like Product Tours for onboarding, Banners for announcements, and Surveys for feedback. These can be useful, but many of them are locked behind the "Proactive Support Plus" add-on, which just adds another line item to your already complicated bill.

Understanding Intercom’s AI: Fin AI Agent

Intercom has gone all-in on AI, with its Fin AI Agent leading the charge. Fin is built to handle support conversations from start to finish, but it’s important to look past the marketing hype to see how it works in the real world.

How the Fin AI Agent works

Fin is an AI chatbot that uses large language models (LLMs). It scans your knowledge base and other Intercom data to give automated answers, do things like tag tickets, and hopefully resolve issues without needing a person. You can turn it on across any channel connected to your inbox, from live chat to email.

A demonstration of the Intercom AI chatbot in action.
A demonstration of the Intercom AI chatbot in action.

Key capabilities and limitations

The marketing for Fin is full of impressive stats about resolution rates and happy customers. And for simple, straightforward questions, it can work pretty well. But many teams eventually bump into a few key problems.

First, Fin can feel like a "black box." It’s hard to tweak its behavior, change its tone of voice, or set up specific rules for when it should hand off a conversation to a human. You’re mostly stuck with the way Intercom decided it should work. This is a huge difference from a tool like eesel AI, which gives you a fully customizable prompt editor and workflow builder. With eesel AI, you get complete control over your AI’s personality, the exact documents it’s allowed to learn from, and the specific actions it can take. It ends up feeling like a true extension of your team, not a generic robot.

This video introduces Fin, Intercom's AI agent designed to resolve complex customer service queries with high-quality, automated answers.

Second, launching a customer-facing AI without being able to properly test it is a massive gamble. One bad automated experience can sour a customer relationship for good. This is where eesel AI’s simulation mode is a huge advantage. Before your AI ever talks to a real customer, you can run it against thousands of your past support tickets in a safe environment. This gives you a real forecast of how well it will perform and how much it will save you, letting you fine-tune everything and launch with confidence. While Intercom’s method can feel a bit like "flip the switch and hope for the best," eesel AI is all about simulating, refining, and then deploying.

Pricing explained (and why it gets complicated)

Reddit
If there’s one thing that consistently frustrates Intercom users, it’s the pricing.

The model has so many different layers that it’s tough to predict your monthly bill, which makes budgeting a nightmare.

Breaking down the costs

Your total Intercom cost isn’t a simple subscription. It’s a mix of a few different things:

  • Seat-Based Plans: You pay a monthly fee for each "Full seat," meaning anyone who needs full access to the platform.

  • AI Resolution Fees: This is the big one. Intercom charges $0.99 for every single conversation resolved by the Fin AI Agent. If you have a decent amount of support traffic, this fee can spiral out of control, fast.

  • Mandatory Add-Ons: A lot of features you might assume are included, like proactive messaging (Product Tours, Surveys) or unlimited AI help for your agents (Copilot), are actually paid add-ons.

  • Usage-Based Channels: Want to use channels like WhatsApp or SMS? That’ll be extra, based on how much you use them.

Here’s a simplified table based on their pricing page:

PlanAnnual Seat PriceFin AI Agent CostKey Features & Add-ons
Essential$29/seat/mo$0.99 per resolutionMessenger, Shared Inbox, Help Center
Advanced$85/seat/mo$0.99 per resolutionMultiple Inboxes, Workflows, 20 Lite seats
Expert$132/seat/mo$0.99 per resolutionSSO, HIPAA support, 50 Lite seats
Add-onsProactive Support ($99/mo), Copilot ($29/agent/mo)

The hidden costs and unpredictability

The biggest problem here is the per-resolution model. It creates a weird situation where the more tickets your AI successfully deflects, the more you have to pay. A busy month or an unexpected spike in questions can lead to a bill that makes your finance team’s eyes water. It makes forecasting your support costs almost impossible.

This is where eesel AI’s pricing is a breath of fresh air. Our plans come with a generous monthly bucket of AI interactions, and we never charge you per resolution. Your bill is the same every month, so it’s completely predictable, no matter how many tickets your AI handles. You can scale your automation without constantly worrying about the cost. Plus, with flexible month-to-month plans, you aren’t locked into a long contract.

Is Intercom the right choice for you?

Intercom is a well-known, established platform with a ton of features. If you’re a new company looking for one tool to do a bit of everything and you aren’t too worried about the budget, it might seem like a good option.

But the downsides are pretty significant. It’s expensive, with a pricing model that almost feels like it punishes you for growing. It locks you into their ecosystem, making it hard to use other best-in-class tools. And its AI just doesn’t offer the control, testing abilities, or broad knowledge integrations that you can get with more modern, specialized AI platforms.

Before you make a call, ask yourself a few questions:

  • Can my budget handle unpredictable AI fees that go up every time we get busier?

  • Is our company knowledge scattered across places like Confluence, Google Docs, or old Slack threads, and not just in one help center?

  • How important is it for me to test and see exactly how an AI will perform before it talks to my customers?

If you found yourself nodding along to any of these, you’d probably be better off with a more flexible setup. eesel AI is built for teams who want powerful AI without all the baggage. It improves the tools you already love, like Intercom or Zendesk, instead of forcing you to migrate to a whole new, expensive platform.

A smarter way to automate your support

Intercom is a powerful platform, no doubt. But for many teams, its high costs, confusing pricing, and "black box" AI are major roadblocks. The pay-per-resolution model for Fin creates budget anxiety and penalizes you for being good at automated support.

But you don’t have to throw out your entire support setup to get great AI. You can simply add a dedicated AI layer on top of your existing helpdesk that gives you all the control you need.

eesel AI connects seamlessly with tools like Intercom, Zendesk, and Freshdesk to give you a fully controllable, testable, and predictably priced AI support agent. It pulls in knowledge from wherever it lives and can be up and running in minutes, not months.

Ready to see how it works? Start your free eesel AI trial today.

Frequently asked questions

Intercom’s pricing combines seat-based fees with a significant $0.99 charge for every conversation resolved by its Fin AI Agent. This per-resolution model, along with mandatory add-ons for many features, can make budgeting unpredictable and costs easily escalate with increased usage.

The Fin AI Agent can feel like a "black box," offering limited control over its behavior, tone, or specific hand-off rules to human agents. Furthermore, Intercom lacks a robust simulation mode, meaning you cannot thoroughly test its performance against past support tickets before deploying it live.

While Intercom includes its own Help Center for knowledge management, it primarily focuses on content created within its platform. Incorporating knowledge from external sources like Google Docs or Confluence for its AI often requires manual effort, which can be time-consuming.

The core features of Intercom encompass the Messenger for real-time customer chat, a Shared Inbox for managing conversations across various channels, a Help Center for self-service articles, and the Fin AI Agent for automated support. These components are designed to offer an all-in-one communication solution.

Intercom’s approach to deploying its Fin AI Agent does not include a comprehensive simulation mode for thoroughly testing its performance against historical support data. This means that launching the AI can feel like a "flip the switch and hope for the best" scenario.

The Intercom Messenger is the customer-facing chat interface, while the Shared Inbox centralizes messages from your website, app, email, and social media into one stream. It provides a basic ticketing system and routing rules to help teams efficiently manage and respond to customer queries from a single platform.

Share this post

Kenneth undefined

Article by

Kenneth Pangan

Writer and marketer for over ten years, Kenneth Pangan splits his time between history, politics, and art with plenty of interruptions from his dogs demanding attention.