A practical guide to using Gorgias AI to classify how to questions vs account specific issues

Kenneth Pangan
Written by

Kenneth Pangan

Stanley Nicholas
Reviewed by

Stanley Nicholas

Last edited October 29, 2025

Expert Verified

Running an e-commerce store means your support inbox is a constant mix of the same old questions and unique problems that need a human touch. You've likely considered using an AI tool like Gorgias to handle the simple stuff, freeing up your team for the trickier issues.

But there's a catch. How does an AI tell the difference between a general "how-to" question and a specific problem tied to a customer's account? If it gets this wrong, you end up with annoyed customers, clunky responses, and an even bigger headache for your support team.

Let's dig into how Gorgias AI is used to classify how to questions vs account specific issues, where it hits a wall, and what a more flexible approach to support automation looks like.

What is Gorgias AI?

Gorgias is a well-known help desk built for e-commerce brands, especially if you're using Shopify or BigCommerce. It basically pulls all your customer emails, chats, and social media messages into one tidy dashboard.

The AI part of it, called AI Agent, is meant to automate support by figuring out what customers are asking and firing back automated answers. Its biggest plus is how tightly it connects with e-commerce platforms. It can pull order details and customer history right into a ticket, which gives both your human agents and the AI a bit more context to work with.

A screenshot of the Gorgias dashboard, a key tool when Gorgias AI is used to classify how to questions vs account specific issues.
A screenshot of the Gorgias dashboard, a key tool when Gorgias AI is used to classify how to questions vs account specific issues.

How Gorgias AI classifies tickets

So, how does Gorgias actually sort through incoming tickets? It uses a feature called "Intents," which is a fancy way of saying it tries to guess the customer's goal.

Understanding Gorgias's "Intents" feature

When a message comes in, the AI scans it and slaps on a label like "shipping/status" for "Where's my package?" or "return/request" for, well, return requests. Gorgias has a set list of these intents it can recognize. If a customer's question fits neatly into one of these boxes, the AI can kick off an automated response.

This system is pretty good for simple, "how-to" questions where the answer is always the same for everyone. Think of things like:

  • What's your return policy?

  • Do you ship internationally?

  • How do I take care of this product?

These are easy to deflect to a help center article or a pre-written answer.

A view of the Gorgias
A view of the Gorgias "Intents" feature, which helps show how Gorgias AI is used to classify how to questions vs account specific issues.

Setting up automations with "When, If, Then"

To make the AI do something with these intents, you have to give it specific instructions using a feature called "Guidance." It's based on a simple "When, If, Then" formula, which they explain in their blog.

For instance:

  • When: A customer asks about a return (the "return/request" intent is triggered).

  • If: Their order is less than 30 days old.

  • Then: Send them the link to the returns portal.

It sounds simple enough, but the catch is that you have to manually map out every single one of these scenarios yourself. This can turn into a lot of work, both to set up and to keep updated as your business changes. You're also stuck with the conditions Gorgias gives you, which can feel a bit restrictive.

The challenge: How-to questions vs account-specific issues

Answering basic FAQs is one thing, but the real test for any support AI is dealing with the messy, high-stakes questions that are unique to each customer. This is where things can get tricky with Gorgias AI, and you'll see people on Reddit and other forums talking about these exact frustrations.

Why account-specific issues are different

Account-specific problems need more than just a canned answer; they often require the AI to take action and check live data from different places. We're talking about questions like:

  • "My discount code isn't working for the stuff in my cart."

  • "Can you change the shipping address for my last order?"

  • "Why did my subscription renewal cost more this month?"

To answer these, the AI can't just quote your help center. It needs to actually connect to your e-commerce platform, your subscription tool, or your shipping provider, look up specific info, and maybe even make a change. This is where the fixed "Intents" and basic "If/Then" rules just don't cut it. Often, the AI either spits out a generic "we'll look into it" response or just gives up and passes the ticket to a human, which kind of defeats the whole point.

Where Gorgias AI struggles

So, where does Gorgias AI tend to stumble?

First, its ability to take custom actions is pretty rigid. While it can do things like edit an order, connecting it to other systems outside of its main integrations can be a real project. There’s no simple way to set up custom actions that pull live data from any API you might be using.

Second, the tone can feel robotic. Without access to rich, real-time information, the AI's answers often sound generic. You can customize the persona, but it's limited. This is when you start hearing customers say, "I just want to talk to a person."

Finally, it's slow to learn. Gorgias AI learns from your help docs and the "Guidance" rules you write. It doesn't automatically pick up on how your team has solved similar problems in the past, which means you're stuck constantly updating its instructions by hand.

A more flexible alternative: eesel AI

This is exactly the kind of gap that a tool like eesel AI is designed to fill. Instead of boxing you into a set menu of intents and actions, it gives you a flexible workflow engine that can plug right into your current helpdesk, including Gorgias.

Here’s a quick look at how it handles those challenges differently:

  • It learns from everything. eesel AI doesn't just scan your help center. It learns from your team's past ticket responses, Google Docs, Confluence, and over 100 other sources. This helps it understand your brand's voice and solutions from the get-go.

  • You can build genuinely custom actions. A simple prompt editor lets you tell the AI exactly how to behave and what to do. Need it to check live order info or query your shipping partner's database? You can set that up in a few minutes, usually without needing a developer.

  • You can test it safely. Worried about letting an AI loose on your customers? eesel AI has a simulation mode that lets you test your setup on thousands of your own past tickets. You can see exactly how it would have responded and get a solid prediction of your automation rate before it ever talks to a real customer.

This level of control means you can automate not just the simple stuff, but also a good chunk of those complex, account-specific issues with confidence.

A screenshot of the eesel AI workflow screen, which is a key component when Gorgias AI is used to classify how to questions vs account specific issues.::
A screenshot of the eesel AI workflow screen, which is a key component when Gorgias AI is used to classify how to questions vs account specific issues.

Understanding Gorgias AI pricing

When you're looking at AI tools, the price tag is obviously a huge factor. A confusing pricing model can quickly eat up any money you thought you were saving. Gorgias's pricing is based on the number of tickets you handle, but the AI features are an extra cost that can be tough to predict.

Gorgias's pricing model breakdown

Based on their official pricing page, here's a quick look at their plans:

PlanMonthly Price (Annual)Included Helpdesk TicketsOverage Cost (per 100 tickets)
Starter$10/mo50$40
Basic$50/mo300$40
Pro$300/mo2,000$36
Advanced$750/mo5,000$36

But here's the key detail: the AI Agent is billed separately. You pay for each "automated interaction," which is usually around $0.90. On top of that, every AI interaction also uses up one of your plan's "billable helpdesk tickets."

The problem with per-resolution pricing

This pay-per-resolution model can cause a couple of headaches.

For one, it leads to unpredictable bills. Your costs are tied directly to how many tickets come in and how many the AI handles. During a busy season like Black Friday, you could be in for a nasty surprise.

Reddit
It's a common complaint that the AI credits 'burn way too fast'.

You're also essentially paying twice. You get charged for the AI to resolve the ticket, and then that same ticket counts against your monthly plan limit. This can make the real cost of automation a lot higher than you might think.

Pro Tip
Before you sign up for a plan like this, take your average monthly ticket volume and multiply it by the per-resolution fee. That'll give you a ballpark estimate of your monthly AI spend, which you'll need to add to your base subscription cost.

By contrast, platforms like eesel AI use clear, predictable pricing. You get a generous number of monthly AI interactions (where one interaction is a single AI reply or action), so your costs don't balloon when you get busy. You can also start with a month-to-month plan, giving you more flexibility.

Gorgias AI vs. eesel AI: A direct comparison

When you're trying to pick an AI for your help desk, it really comes down to control, flexibility, and clear pricing. Gorgias offers a system that's built right in, but its rigidity can be a real hurdle.

Here’s a quick side-by-side look:

FeatureGorgias AIeesel AI
SetupYou have to manually write "Guidance" rules for every single thing you want to automate. It can take a while.You can get it up and running in minutes. It learns from your past tickets automatically.
FlexibilityYou're stuck with their preset intents and a strict "When, If, Then" system.You get a flexible prompt editor to fully control the AI's tone, personality, and what it can do.
IntegrationsGreat for Shopify, but connecting to other tools often requires a developer.Has over 100 one-click integrations and lets you build your own connections to any API.
TestingNo test environment. You have to try out your automations on live customers.Includes a simulation mode to test on your old tickets, so you can see how it performs without any risk.
PricingCan be confusing. You pay per AI resolution on top of your base ticket plan, which makes costs hard to predict.Straightforward and predictable. Plans are based on a set number of AI interactions, with no hidden fees.

Move beyond basic classification to true automation

Look, Gorgias AI is a decent starting point for e-commerce brands wanting to try out support automation. Its "Intents" feature does a fine job of sorting common "how-to" questions and sending out canned responses.

But the toughest part of customer support isn't the easy questions, it's handling the complicated, account-specific problems efficiently. That's where a rigid system with a lot of manual setup and confusing pricing can hold you back. If you want to automate a meaningful amount of your support volume without frustrating your customers, you'll need something more flexible and transparent.

A tool like eesel AI lets you go beyond just sorting tickets. By learning from all your company knowledge and past conversations, and by letting you build any custom action you need, it helps you create an AI assistant that can actually resolve nuanced problems. This frees up your team to focus on the conversations that really matter.

Curious to see what a more customizable AI can do for your team? Get started with eesel AI for free.

Frequently asked questions

Gorgias AI uses an "Intents" feature to scan incoming messages and assign predefined labels like "shipping/status" or "return/request". If a customer's query matches one of these intents, the AI can trigger an automated response, which is generally effective for straightforward "how-to" questions.

Gorgias AI struggles with account-specific issues because they often require custom actions, real-time data lookups from various systems, or making specific changes. Its fixed "Intents" and basic "When, If, Then" rules lack the flexibility needed to handle these nuanced and unique customer problems.

You can customize Gorgias AI using "Guidance" based on a "When, If, Then" formula to map specific scenarios to intents. However, this process requires significant manual setup for every scenario and is constrained by Gorgias's predefined conditions and rigid integration capabilities.

Accurate classification is crucial to prevent customer frustration from receiving generic answers to complex, account-specific problems and to free up human agents from repetitive "how-to" queries. It ensures that customers receive appropriate and efficient support, improving satisfaction and operational efficiency.

Gorgias's pay-per-resolution model for its AI Agent, where each automated interaction also counts as a billable helpdesk ticket, can lead to unpredictable and potentially high costs. This dual-charging structure can make it expensive to extensively leverage the AI for classification and automation, especially during peak periods.

Gorgias AI primarily learns from your help documentation and manually written "Guidance" rules. It does not automatically learn from your team's historical ticket responses or past solutions, necessitating constant manual updates and making it slower to adapt to evolving support needs.

eesel AI offers greater flexibility by learning from all company knowledge, including past ticket responses, and provides a customizable workflow engine for building truly custom actions. This allows for deeper automation of complex, account-specific issues without being confined by rigid intents, and includes a simulation mode for risk-free testing.

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.