A complete guide to Fin AI follow up: Features, pricing, and limitations

Let's be honest, managing customer follow-ups is a grind. It’s one of those tasks that’s absolutely essential for keeping customers happy, but it’s also repetitive and eats up a ton of time. Do it well, and your customers feel like you've got their back. Drop the ball, and you risk losing them for good.

This is where AI has stepped in, promising to automate the whole process. Intercom’s Fin is one of the bigger names in this space, designed to jump into customer conversations, solve problems, and keep everything on track without a human needing to babysit it.

But how good is it when it comes to actual follow-ups? This guide will give you the real story on the Fin AI follow up system. We’ll look at its features, walk through the setup, break down the pricing, and point out a few key limitations you need to know about before you go all-in.

What is Intercom's Fin AI follow up system?

Think of Fin AI as Intercom's in-house AI agent for your support team. Its main purpose is to answer customer questions automatically, no matter where they come from, your website's live chat, email, or social media. It's meant to be the first line of defense, handling common questions so your human agents can focus on the trickier stuff.

Under the hood, Fin runs on its "AI Engine," which is just a fancy way of saying it reads a customer's question, digs through your knowledge base to find the right info, and then tries to write a clear answer.

It’s built to handle more than just simple questions. It can manage conversations that require a few back-and-forth steps to get to a resolution. This is where its follow-up features really kick in, using a mix of automated workflows and more complex procedures to guide a customer from start to finish.

How to manage a Fin AI follow up with workflows

The primary tool you'll use to control a Fin AI follow up is Intercom’s “Workflows.” This is where you map out the logic that tells Fin what to do in different scenarios. The key piece of the puzzle is the "Let Fin answer" step, which you can just drop into any workflow to let the AI take over the conversation.

A screenshot of the Intercom Workflows UI, showing how a Fin AI follow up can be configured.

Once Fin is part of a workflow, you get a decent amount of say over how it behaves:

• Targeting your audience: You can use branching logic to give different types of customers completely different experiences. For instance, you could set it up so a paying customer gets routed to a human agent much faster, while a free user is pointed toward self-service articles first.

• Controlling the channels: You get to pick and choose exactly where Fin is active. Maybe you want it on your website chat but not touching your support emails. That's up to you.

• Setting handover rules: You decide when it's time for a human to step in. If a customer gives negative feedback or flat-out asks for a person, you can build a rule that immediately passes the conversation to the right team.

• Closing out conversations: You can have Fin automatically ask for a CSAT score after it believes an issue is solved. You can also tell it to close out conversations if a customer stops responding, which helps keep your team's inbox from getting cluttered.

While Intercom's workflows let you tweak almost everything, getting that initial setup right can be a project in itself. For teams that want to get started without a massive upfront investment of time, other tools can simplify things. For example, eesel AI can go live in minutes by connecting directly to your helpdesk. It also has a really useful simulation mode where you can test the AI on thousands of your past tickets to see how it would have performed. This lets you get a solid performance forecast before you ever turn it on for live customers.

Using procedures and tasks for complex follow-ups

But what about follow-ups that aren't just simple questions? For those trickier, multi-step problems, Fin has a feature called "Procedures." These are designed for processes that involve a bit more business logic, like checking specific conditions or grabbing data from other tools.

Here’s a quick look at how Procedures work:

• Plain English instructions: You can write out the steps for a procedure just like you would for a new team member. You can even copy and paste your existing SOPs or get the AI to draft a procedure from a quick outline.

• Strict rules: Even though the instructions are written in natural language, you can add some hard if/else logic to make sure Fin follows the rules every time. For example, if an order is less than 30 days old, then start the refund process; else, tell the customer they’re past the return window.

• Data connectors: This is how Fin talks to your other software. You can hook it up to systems like Shopify or Stripe to pull live information, like an order's shipping status or a customer's subscription plan, and use that info to decide what to do next.

A classic example is a refund request. A customer asks for their money back, which kicks off the "Refund Procedure." Fin then uses a data connector to check the order date in your system. Based on the 30-day policy you set, it either moves forward with the refund or explains to the customer why it can't.

Be warned, though, building and testing these Procedures can get technical pretty quickly. "Natural language" sounds easy, but making sure the logic is perfect takes careful setup and a lot of testing. You're also stuck with the data connectors Intercom offers, and getting a custom integration built can be a major headache.

Custom actions are the key to unlocking real automation, but you shouldn't need a developer on standby to set them up. With eesel AI, you get a fully customizable workflow engine that’s built to be self-serve. You can create your own API actions to pull information from any system and feed the AI knowledge from all your sources, from Google Docs to past tickets, without being locked into a limited set of connectors.

Fin AI pricing: What a follow-up will cost you

Okay, let's talk about the part everyone's really wondering about: the price. Fin’s pricing model is a bit different from your standard monthly subscription, and it can seriously affect your budget if you're not careful.

Fin’s pricing is based on a per-resolution model. Here’s how it works:

• The fee: Fin costs $0.99 per resolution.

• What counts as a "resolution": Intercom counts a "resolution" anytime Fin gives an answer and the customer either says it was helpful ("confirmed resolution") or just leaves the conversation ("assumed resolution"). You get charged once per conversation, even if Fin answers several questions within it.

• Helpdesk costs: That $0.99 fee is in addition to your helpdesk subscription. If you're using the whole Intercom package, you're paying the per-resolution fee plus your seat price, which starts at around $29 per seat each month.

• Minimums: If you use Fin with a different helpdesk like Zendesk or Salesforce, you have to pay for at least 50 resolutions per month.

The biggest headache here is the lack of predictability. A per-resolution model means your bill can jump around a lot from month to month. If you have a busy season, launch a new product, or run into a bug that floods your support channels, your AI bill could be much higher than you expected. This can put you in an odd spot where you're almost afraid to automate too much because of the potential cost.

This pricing model isn't unique to Fin, but it can be a nightmare for anyone who needs a predictable budget. For teams that value cost certainty, there are alternatives with more straightforward pricing. For instance, eesel AI offers transparent, predictable pricing based on your overall interaction volume, not how many tickets the AI closes. This means your bill doesn't automatically spike with your ticket volume, letting you automate with peace of mind.

Feature	Cost (Standalone Helpdesk)	Cost (with Intercom Suite)
Per Resolution	$0.99	$0.99
Minimum Monthly	50 resolutions	N/A
Helpdesk Seat	N/A (Uses your existing desk)	Starts at $29/seat/month
Billing Model	Unpredictable (scales with volume)	Unpredictable (scales with volume)

The biggest headache here is the lack of predictability. A per-resolution model means your bill can jump around a lot from month to month. If you have a busy season, launch a new product, or run into a bug that floods your support channels, your AI bill could be much higher than you expected. This can put you in an odd spot where you're almost afraid to automate too much because of the potential cost.

This pricing model isn't unique to Fin, but it can be a nightmare for anyone who needs a predictable budget. For teams that value cost certainty, there are alternatives with more straightforward pricing. For instance, eesel AI offers transparent, predictable pricing based on your overall interaction volume, not how many tickets the AI closes. This means your bill doesn't automatically spike with your ticket volume, letting you automate with peace of mind.

Is a Fin AI follow up the right move for you?

At the end of the day, is Fin the best tool for handling your follow-ups? It really depends on your situation. Fin AI is a powerful and tightly integrated system, and if you're already living and breathing in the Intercom ecosystem, its Workflows and Procedures give you a massive amount of control.

But that power comes with some serious trade-offs. The setup can be complicated and chew up a lot of time, especially for the more advanced Procedures. More importantly, the per-resolution pricing can make budgeting a constant challenge, which is a big deal when you're trying to scale your support. It also ties you even more tightly to one platform, which might limit your options later on.

So it really boils down to this: are you looking for a powerful system and are willing to deal with the potential cost and complexity, or would a more flexible, easy-to-use, and predictably priced tool be a better fit for your team?

If you want the benefits of AI automation without the surprise bills and complicated setup, it might be worth checking out a more flexible option. eesel AI is a simple, self-serve platform that connects to your current helpdesk, pulls in knowledge from all your sources instantly, and offers flat, predictable pricing. Start a free trial today and see for yourself how fast you can get your support automated with total confidence.