Chat for website is software that puts a message window on your site so visitors can ask a question without leaving the page. A widget collects the message, an inbox routes it, and a person, a bot or an AI agent answers. For a store, the deciding factor is where the chat lands and whether the answerer sees the order.
What is chat for website?
Chat for website, also sold as live chat software, has three parts: the widget, the message window on the page; the inbox, where the conversation lands and is routed; and the answerer, which is a person, a rule-based bot or an AI agent.
The widget is the visible part; the inbox and the answerer decide what the visitor gets. For the full picture of how this fits an online store, start with our hub.
How does website chat software work?
A snippet or app embed adds the widget to your pages, the visitor’s message lands in an inbox, routing rules decide who sees it, and someone answers in real time or later. The same mechanics sit behind most live chat programs, whatever they are called.
Install. Zendesk’s messaging Web Widget is code you embed in your website or help centre. For Shopify setup, see our guide to Shopify chat.
Landing. Where the message goes differs by tool. In Gorgias, each chat conversation becomes a ticket in the helpdesk. In Zendesk messaging, a ticket is created when a customer asks for a live agent. In Shopify Inbox, staff reply from the desktop or mobile apps.
Routing and answering. Someone has to decide who answers first (a bot, an AI agent or a person) and what happens when nobody is online. That choice, more than the widget, is what the visitor experiences.
What types of website chat are there?
There are five types of website chat: the live chat widget, the rule-based chatbot, the AI agent, help desk messaging and in-app chat. They differ in who answers, where the conversation lands and whether the answerer can act on an order.
| Type | Who answers | Where it lands | Can it act on an order? | Typical fit |
|---|---|---|---|---|
| Live chat widget | People, in real time | The chat tool’s own console | Only if a person does it by hand | One small team answering pre-sale questions |
| Rule-based chatbot | Scripted flows and FAQs | The bot’s flow, then a hand-off | No, it follows the script | Simple, repeated questions |
| AI agent | An AI that reads data and follows policies | Your help desk or chat tool | Yes: it reads live data, acts in connected systems and replies, with hand-offs to your team set by your policies | Order questions at volume |
| Help desk messaging | People and AI, in real time or later | A persistent thread inside the help desk | Through the help desk’s integrations | Stores that run several channels |
| In-app chat | People, bots or AI | A messenger inside a mobile app, installed through an SDK | Depends on the tool behind it | Stores with their own app |
Zendesk distinguishes the two: live chat is real-time and session based, per Zendesk, while messaging conversations can be picked back up later. Messaging is enabled by default on new Zendesk instances.
We do not redefine chatbots and AI agents here: that lives in our chatbot comparison. For the bot route, see website chatbots.
A website with chat support is simply a site where visitors can open a chat to reach the business, answered by people, a bot or an AI agent.
How common is live chat on websites?
Live chat is everywhere, and the tools are plentiful. Wappalyzer tracks 376 live chat technologies in its 2026 ranking and offers contact data for 3,200,000 websites using them.
The most common one is not a chat tool in the classic sense. In the same Wappalyzer ranking, WhatsApp Business Chat leads with 1,100,000 sites tracked (35% share). The 2026 ranking also includes Joinchat (200,000, 6%), Tawk.to (180,000, 6%), Zendesk (170,000, 5%), Pusher (120,000, 4%), Intercom (110,000, 4%) and Tidio (98,000, 3%).
Market size is harder to pin down, because estimates differ by definition. IMARC Group puts the global live chat software market at USD 1,227.6 million in 2025, forecast to reach USD 2,185.5 million by 2034 (6.42% CAGR, 2026 to 2034). It names customer service live chat systems as the largest segment and North America as the largest region. The Business Research Company estimates $1.34 billion in 2025 and $1.49 billion in 2026 (11.2% CAGR).
Our reading: with 376 tools, and a WhatsApp button as the most common “live chat” on the web, the widget is a commodity. Pick on where the chat lands and what the answerer can see.
What jobs does website chat do for a store?
Website chat does three jobs for a store: it answers pre-sale questions about products, sizing and shipping, it answers order questions after purchase, and it lets customers ask for a change such as a return, a cancellation or a new address. Each job needs something different from the answerer.
Pre-sale questions need product and policy knowledge. Order questions need the order. Changes need the ability to act in the shop system, not only to reply.
What can go wrong with website chat?
Four things go wrong most often with website chat: a widget with no inbox behind it, a bot that dead-ends, no plan for offline hours, and an answerer who cannot see the order.
A widget without an inbox means messages pile up in a console nobody owns. A bot dead end leaves the visitor with no route to a person. With no offline plan, a visitor types into a window and nothing happens. And an answerer without order access can only ask the customer to repeat what the shop already knows.
What should you look for in customer support chat software?
Judge customer support chat software on the nine questions below before you judge how the widget looks, starting with how a person takes over and what the answerer can do. The same questions apply to online chat support software and live chat support software under any name.
| Criterion | Question to ask the vendor | Why it matters for a store |
|---|---|---|
| Where conversations land | Do chats go to your own console or into your help desk ticket queue? | One queue means one customer history |
| Order context | Can the answerer see the customer’s orders next to the chat? | A where-is-my-order answer lives in the order, not the chat |
| Actions | Can refunds, cancellations or address changes happen without leaving the conversation? | It saves a swivel-chair step per ticket |
| Conversation history | Is it session-based or persistent? Can the customer come back later? | A customer who comes back should not start again |
| Hand-off | Does a person get the full transcript and context? | Customers should not repeat themselves |
| AI transparency | Does an AI answer show which source or policy it used? | You can check and correct it |
| Install and page speed | Is it an app embed or a snippet? Have you tested before and after? | A widget loads on every page |
| Channels | Is there one thread across web chat, email, WhatsApp, social and in-app? | Customers switch channels mid-issue |
| Offline behaviour | What does a visitor see when nobody is online? | A visitor needs to know what happens when nobody answers |
The actions row is not hypothetical. Zendesk’s Shopify integration lets agents view Shopify order information and process Shopify order refunds and cancellations without leaving the ticket.
Should website chat land in a standalone tool or your help desk?
Land it in your help desk once more than one person answers, or once chat is not your only channel. A standalone chat console splits the customer’s history from email and social, and the person answering may not see the order.
That is why the landing point decides who can see the order. Help desk integrations put order data beside the conversation: Zendesk’s Shopify integration shows order information in the ticket, and Gorgias turns each chat into a ticket in the helpdesk. If you run either, our Zendesk integration and Gorgias integration pages show how Engaige connects, and our Zendesk chat guide covers the chat setup.
A standalone tool is fine in one case: a single person, low volume and pre-sale questions only. Once you have a team or several channels, move chat into the queue where the rest of your support lives.
What is in-app chat support?
In-app chat support is the same messenger inside a mobile app, installed through an SDK. Intercom offers mobile SDKs for iOS, Android and React Native, and Zendesk messaging works through a mobile SDK for Android or iOS apps. If your store has its own app, the in-app thread should land in the same queue as web chat.
Where does Engaige fit?
Engaige (that’s us) is the e-commerce AI support agent that resolves the conversations your website chat feeds into your help desk. Keep your helpdesk. Replace the AI. Your chat tool and help desk stay; what changes is who answers inside them. Engaige AI learns from your conversations, configures your support agent and shows which policies it used. You set policy by talking to Engaige AI in plain language, like briefing a colleague.
You go live in Agent Assist in hours: Engaige drafts replies for your team. Engaige AI compares what your team said with what the agent suggested, learns from the difference and improves the agent, under your supervision: you approve the changes.
As your confidence grows you switch ticket types to fully autonomous; brands that move quickly get there in about a week, and Engaige automates up to 90% of supported tickets. It keeps learning once autonomous, from emerging trends and the tickets where the agent hands back to your team.
Engaige is text-first by design: email, chat, WhatsApp, SMS and social DMs, no voice AI. A live call gives an AI seconds to reason; text gives it time to read the order, check your policy and act, so resolution quality is higher.
Engaige works on the conversations already in your help desk or Crisp. The action happens in the connected system first, then it replies and closes, and hand-offs to your team follow your policies, with context.
On Shopify it combines live Admin API data with the content of your storefront.
At MR MARVIS, a Shopify Plus slow-fashion brand, Engaige handles over 120,000 questions a year, resolving over 60% of the conversations it’s involved in; Otrium resolves 60% of 120,000 annual tickets end to end. See MR MARVIS for the full story.
With Engaige, we are able to automatically resolve 60% of the 120,000 support tickets that we receive annually.
Engaige also gives product advice inside the storefront chat widget you run, usually the help centre widget. The full overview is in our hub, and you can see it on your own tickets in a demo.
Two sibling guides own the neighbouring questions. Website chatbots covers the bot route, and chat benefits covers why stores offer chat. For the AI route, see AI agents for customer service and online printing automation for one industry’s view.
Frequently asked questions
What is a live assistant?
On a website, a live assistant is a person, or increasingly an AI agent, answering visitors in real time through chat.
Is live chat software the same as a chatbot?
No. Live chat routes visitors to people, a chatbot answers from scripts, and an AI agent can read data and act. The difference is covered in our chatbot comparison.
What is the difference between live chat and messaging?
Per Zendesk, live chat is real-time and session based, while messaging can be picked back up later without losing context or history.
How many websites use live chat?
Wappalyzer tracks 376 live chat technologies and 3,200,000 websites using them in its 2026 ranking. WhatsApp Business Chat is the most common, at 1,100,000 sites.
Should my chat widget connect to my help desk?
Yes, once more than one person answers or you run several channels. You get one queue, one history and the order beside the chat.
Is Engaige live chat software?
Engaige is the AI support agent behind your website chat: it answers in your help desk and gives product advice in the storefront chat widget, usually the help centre widget. You keep your chat tool and help desk.
Why is Engaige text-first and not voice?
Engaige is text-first by design: email, chat, WhatsApp, SMS and social DMs, no voice AI. A live call gives an AI seconds to reason; text gives it time to read the order, check your policy and act, so resolution quality is higher.
Can an AI agent answer website chat for an online store?
Yes. For the bot route, see website chatbots and the chatbot comparison. Engaige takes the action in the connected system first, then replies and closes; hand-offs follow your policies. See WISMO automation or book a demo.
