Magento 2 AI Chatbot – eChat

Not a demo claim: over three months, eChat handled 3,766 conversations across dozens of live Magento stores, with no human agent behind any of them. Every number on this page comes from that data.

How It Works

eChat doesn't just answer questions. It actively helps customers find and choose the right products, turning hesitation into purchases. Across live Magento stores running eChat, 38% of all conversations start with a product question: shoppers use chat to decide what to buy, not just to complain.

The most common support questions are the simplest ones. eChat handles them automatically so your team can focus on issues that actually need a human. In production it fully resolved about 50% of conversations on its own, and fully or partially resolved about 85%. Only 0.66% of shoppers asked for a live person, and honesty is the likely reason: the widget says it's an AI, so people just ask their question. We never pass the bot off as a human.

Most chatbots only generate text. eChat can use tools: it connects to your Magento store to look up real data, pull documents, and perform actions on behalf of the customer.

Four tools ship by default: create a ticket, look up order status, get tracking, and pull an invoice. Each one connects eChat to a specific part of your store, turning the chat into a real self-service hub.

A handoff is usually a limit on authority, not on understanding. eChat can read a return request perfectly well, but it has no right to process a return on your behalf, so it lays out your return terms and explains how to request one.

You draw that line yourself. Escalation scenarios pair a trigger situation with what eChat should do about it: route an unhappy customer to support and end the conversation, open a ticket right away, or hand over your contact details. The user manual covers how scenarios are configured.

When a handoff does happen, nobody is dropped into nowhere. eChat confirms the customer's email, pulls the request text, the conversation context, and the order number out of the chat itself, and hands your team a prepared case instead of a "hi, I have a problem".

eChat indexes all of your Magento content to provide accurate, store-specific answers:

• Products: names, descriptions, attributes, prices, and availability
• Categories: structure, navigation, and category-specific content
• CMS Pages: policies, FAQs, about us, shipping info, and more
• Knowledge Base: articles, guides, and help content from your knowledge base

When you add new products, update prices, or change a return policy, eChat picks up the changes automatically through reindexing. No manual updates, no stale answers.

Your content can stay in one language. eChat detects the shopper's language from their first word and runs the whole conversation in it: in production, shoppers chatted in 30+ languages while the catalogs and knowledge bases behind them stayed in English.

Hallucinations are mostly a reaction to a gap in your content: unable to find the information, the model starts filling it in. Instructions alone don't fix that, so eChat doesn't rely on them. An answer never goes straight to the shopper. A guard layer first checks the facts in it against your knowledge base.

We would rather name the number than pretend it's zero: in production, that layer still lets a false answer through roughly once every thousand conversations. That is why the two mechanisms work as a pair. Whatever the guard kills, and whatever the bot couldn't answer at all, goes into the unanswered-questions queue for you to fix at the source.

Real Questions, Real Answers

Here's what customers can ask eChat, and get accurate, instant answers.

Cents, not dollars. Across our production data, the median conversation (usually three to five question-and-answer pairs) came to $0.028 of model cost. No human agent answers a question for three cents.

What drives the number is not message length but how much information the bot has to read before it can answer. One question is rarely one model call: a hard product question makes eChat search, open a page, see that it doesn't answer, open the next one, and come back to compare. A person researches a topic the same way, except a person isn't billed per token.

The heaviest conversations in our data were product consultations, the ones you least want cut short, and they still stayed within single dollars. That is why we don't recommend capping a conversation by cost. Abuse is a separate problem with a separate answer, and dedicated layers cut off anyone who dumps huge texts into the chat or tries to pull it away from your store's topic.

Across the stores in our production data, over 90% of what the bot knows is the catalog itself. The remaining 10% comes from CMS pages and content the merchant added on purpose. That is why the information most often missing is so unglamorous:

• live stock
• prices and current promos
• store addresses
• shipping terms
• self-service instructions
• sometimes even the company name, because it lives in the logo and nowhere in the data the chat reads

These are not AI blind spots. They are store blind spots that eChat makes visible. Two questions tell you how much preparation you need: what do customers ask about most, and is the answer already in your admin panel? If your product descriptions are complete and detailed, you may need no preparation at all.

When eChat Is Not the Right Fit

Two cases where we would tell you to wait:

• Your store has almost no traffic. A chatbot multiplies conversations that are already happening. It does not create visitors.
• Your answers live in your support team's heads. If the information was never published on the site, the bot will expose that gap in the first week. Content comes first: detailed product descriptions, policy pages, a basic FAQ.

There is also a limit on the role itself. We see eChat as an information helper, not a sales consultant. Selling is emotion, psychology, and the argument for this particular product, and a human is still stronger at that. If you run a live sales chat today, keep it. If you never had one, nobody was selling in this channel anyway, and now at least someone answers product questions at three in the morning.

Your customers have questions at every hour of the day. eChat makes sure they always get answers directly from your store's own content: accurate, helpful, and instant.